{ "generated_at": "2026-03-21T04:14:32.000Z", "latest_snapshot_at": "2026-03-21T04:14:32.000Z", "counts": { "models": 37, "providers": 19, "agents": 14, "changes": 16 }, "models": [ { "id": "openrouter::qwen/qwen3-coder:free", "model_id": "qwen/qwen3-coder:free", "name": "Qwen3 Coder 480B", "source": "openrouter", "provider": "Qwen (Alibaba)", "context_window": 262000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "Massive 480B parameter coding model (35B active MoE). Top free coding model for agentic workflows.", "pricing_note": "Free on OpenRouter, may train on data", "best_for": "Code generation, agentic workflows", "quality_tier": "frontier", "status": "available", "first_seen_at": "2026-03-07T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::qwen/qwen3-next-80b-a3b-instruct:free", "model_id": "qwen/qwen3-next-80b-a3b-instruct:free", "name": "Qwen3-Next-80B-A3B", "source": "openrouter", "provider": "Qwen (Alibaba)", "context_window": 262000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "Next-gen Qwen3 80B (3B active). Fast, no thinking traces, great for RAG.", "pricing_note": "Free on OpenRouter", "best_for": "RAG, instruction following", "quality_tier": "high", "status": "available", "first_seen_at": "2026-02-24T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::stepfun/step-3.5-flash:free", "model_id": "stepfun/step-3.5-flash:free", "name": "Step 3.5 Flash", "source": "openrouter", "provider": "StepFun", "context_window": 256000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": false, "description": "StepFun reasoning model, 196B (11B active MoE). Speed-efficient.", "pricing_note": "Free on OpenRouter", "best_for": "Reasoning, speed-efficient tasks", "quality_tier": "mid", "status": "available", "first_seen_at": "2026-03-03T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::nvidia/nemotron-3-nano-30b-a3b:free", "model_id": "nvidia/nemotron-3-nano-30b-a3b:free", "name": "Nemotron 3 Nano 30B", "source": "openrouter", "provider": "NVIDIA", "context_window": 256000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": false, "description": "NVIDIA 30B (3B active MoE). Highest compute efficiency among free models.", "pricing_note": "Free on OpenRouter", "best_for": "Efficient inference", "quality_tier": "mid", "status": "available", "first_seen_at": "2026-02-28T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::nvidia/nemotron-3-super-120b-a12b:free", "model_id": "nvidia/nemotron-3-super-120b-a12b:free", "name": "Nemotron 3 Super", "source": "openrouter", "provider": "NVIDIA", "context_window": 1000000, "supports_tool_calling": false, "supports_vision": false, "supports_structured_output": false, "description": "NVIDIA open reasoning model with a 1M context window and strong benchmark performance for long-horizon tasks.", "pricing_note": "Free on OpenRouter", "best_for": "Long-context reasoning, agent planning", "quality_tier": "frontier", "status": "available", "first_seen_at": "2026-03-21T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::openrouter/free-router", "model_id": "openrouter/free-router", "name": "OpenRouter Free Router", "source": "openrouter", "provider": "OpenRouter", "context_window": 200000, "supports_tool_calling": true, "supports_vision": true, "supports_structured_output": true, "description": "Auto-selects from available free models. Meta-router for best available free model.", "pricing_note": "Free on OpenRouter", "best_for": "Automatic free model selection", "quality_tier": "mid", "status": "available", "first_seen_at": "2026-02-15T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::arcee-ai/trinity-large-preview:free", "model_id": "arcee-ai/trinity-large-preview:free", "name": "Arcee Trinity Large Preview", "source": "openrouter", "provider": "Arcee AI", "context_window": 131000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "400B (13B active MoE). Trained for Kilo/OpenCode/Cline coding agents.", "pricing_note": "Free on OpenRouter", "best_for": "Tool use, agentic tasks", "quality_tier": "high", "status": "available", "first_seen_at": "2026-02-21T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::arcee-ai/trinity-mini:free", "model_id": "arcee-ai/trinity-mini:free", "name": "Arcee Trinity Mini", "source": "openrouter", "provider": "Arcee AI", "context_window": 131000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "Compact 26B (3B active) version of Trinity. 128 experts.", "pricing_note": "Free on OpenRouter", "best_for": "Lightweight agentic tasks", "quality_tier": "mid", "status": "available", "first_seen_at": "2026-02-21T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::zhipuai/glm-4.5-air:free", "model_id": "zhipuai/glm-4.5-air:free", "name": "GLM-4.5-Air", "source": "openrouter", "provider": "Z.AI (Zhipu)", "context_window": 131000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "Z.AI GLM with thinking + non-thinking modes.", "pricing_note": "Free on OpenRouter", "best_for": "General coding, tool use", "quality_tier": "mid", "status": "available", "first_seen_at": "2025-09-22T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::qwen/qwen3-vl-235b-a22b-thinking:free", "model_id": "qwen/qwen3-vl-235b-a22b-thinking:free", "name": "Qwen3-VL-235B Thinking", "source": "openrouter", "provider": "Qwen (Alibaba)", "context_window": 131000, "supports_tool_calling": true, "supports_vision": true, "supports_structured_output": true, "description": "Multimodal reasoning 235B (22B active). STEM/math with vision.", "pricing_note": "Free on OpenRouter", "best_for": "Multimodal reasoning, STEM", "quality_tier": "high", "status": "available", "first_seen_at": "2026-02-19T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::qwen/qwen3-vl-30b-a3b-thinking:free", "model_id": "qwen/qwen3-vl-30b-a3b-thinking:free", "name": "Qwen3-VL-30B Thinking", "source": "openrouter", "provider": "Qwen (Alibaba)", "context_window": 131000, "supports_tool_calling": true, "supports_vision": true, "supports_structured_output": true, "description": "Lighter multimodal 30B (3B active). Vision + text reasoning.", "pricing_note": "Free on OpenRouter", "best_for": "Vision + text reasoning", "quality_tier": "mid", "status": "available", "first_seen_at": "2026-02-19T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::openai/gpt-oss-120b:free", "model_id": "openai/gpt-oss-120b:free", "name": "GPT-OSS 120B", "source": "openrouter", "provider": "OpenAI", "context_window": 131000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "OpenAI open-weight 117B (5.1B active MoE). Single H100, chain-of-thought.", "pricing_note": "Free on OpenRouter", "best_for": "Reasoning, chain-of-thought", "quality_tier": "high", "status": "available", "first_seen_at": "2026-02-09T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::openai/gpt-oss-20b:free", "model_id": "openai/gpt-oss-20b:free", "name": "GPT-OSS 20B", "source": "openrouter", "provider": "OpenAI", "context_window": 131000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "OpenAI open-weight 21B (3.6B active). Apache 2.0, runs on consumer GPU.", "pricing_note": "Free on OpenRouter", "best_for": "Consumer GPU inference", "quality_tier": "mid", "status": "available", "first_seen_at": "2026-02-04T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::nvidia/nvidia-nemotron-nano-9b-v2:free", "model_id": "nvidia/nvidia-nemotron-nano-9b-v2:free", "name": "Nemotron Nano 9B v2", "source": "openrouter", "provider": "NVIDIA", "context_window": 128000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": false, "description": "Unified reasoning/non-reasoning 9B model.", "pricing_note": "Free on OpenRouter", "best_for": "Quick tasks, lightweight use", "quality_tier": "budget", "status": "available", "first_seen_at": "2025-11-21T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::nvidia/nemotron-nano-12b-2-vl:free", "model_id": "nvidia/nemotron-nano-12b-2-vl:free", "name": "Nemotron Nano 12B VL", "source": "openrouter", "provider": "NVIDIA", "context_window": 128000, "supports_tool_calling": false, "supports_vision": true, "supports_structured_output": false, "description": "Hybrid Transformer-Mamba 12B. OCR/video understanding.", "pricing_note": "Free on OpenRouter", "best_for": "OCR, video understanding", "quality_tier": "mid", "status": "available", "first_seen_at": "2026-01-20T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::meta-llama/llama-3.3-70b-instruct:free", "model_id": "meta-llama/llama-3.3-70b-instruct:free", "name": "Llama 3.3 70B Instruct", "source": "openrouter", "provider": "Meta", "context_window": 128000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "GPT-4 class multilingual 70B model from Meta.", "pricing_note": "Free on OpenRouter", "best_for": "General tasks, multilingual", "quality_tier": "high", "status": "available", "first_seen_at": "2025-09-02T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::google/gemma-3-27b-it:free", "model_id": "google/gemma-3-27b-it:free", "name": "Gemma 3 27B", "source": "openrouter", "provider": "Google", "context_window": 128000, "supports_tool_calling": true, "supports_vision": true, "supports_structured_output": true, "description": "Vision-language 27B, 140+ languages.", "pricing_note": "Free on OpenRouter", "best_for": "Vision, multilingual", "quality_tier": "mid", "status": "available", "first_seen_at": "2025-03-21T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::google/gemma-3-12b-it:free", "model_id": "google/gemma-3-12b-it:free", "name": "Gemma 3 12B", "source": "openrouter", "provider": "Google", "context_window": 128000, "supports_tool_calling": true, "supports_vision": true, "supports_structured_output": true, "description": "Lighter Gemma variant, 12B.", "pricing_note": "Free on OpenRouter", "best_for": "Fast tasks, vision", "quality_tier": "budget", "status": "available", "first_seen_at": "2025-03-21T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::deepseek/deepseek-r1-0528:free", "model_id": "deepseek/deepseek-r1-0528:free", "name": "DeepSeek R1 0528", "source": "openrouter", "provider": "DeepSeek", "context_window": 128000, "supports_tool_calling": false, "supports_vision": false, "supports_structured_output": false, "description": "Open reasoning tokens, o1-class. Strong at math, logic, step-by-step.", "pricing_note": "Free on OpenRouter", "best_for": "Reasoning, math, logic", "quality_tier": "high", "status": "available", "first_seen_at": "2025-06-14T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::moonshotai/kimi-k2:free", "model_id": "moonshotai/kimi-k2:free", "name": "Kimi K2", "source": "openrouter", "provider": "MoonshotAI", "context_window": 128000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "Agentic tool use + code synthesis MoE model.", "pricing_note": "Free on OpenRouter", "best_for": "Tool use, code synthesis", "quality_tier": "high", "status": "available", "first_seen_at": "2025-07-14T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::mistralai/mistral-small-3.1-24b-instruct:free", "model_id": "mistralai/mistral-small-3.1-24b-instruct:free", "name": "Mistral Small 3.1 24B", "source": "openrouter", "provider": "Mistral", "context_window": 96000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "Instruction-tuned 24B, multilingual.", "pricing_note": "Free on OpenRouter", "best_for": "Fast inference, multilingual", "quality_tier": "mid", "status": "available", "first_seen_at": "2025-05-25T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::mistralai/devstral-2512:free", "model_id": "mistralai/devstral-2512:free", "name": "Devstral 2512", "source": "openrouter", "provider": "Mistral", "context_window": 128000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "Coding-focused, agentic model from Mistral.", "pricing_note": "Free on OpenRouter", "best_for": "Coding, development", "quality_tier": "high", "status": "available", "first_seen_at": "2025-12-11T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::allenai/olmo-3.1-32b-think:free", "model_id": "allenai/olmo-3.1-32b-think:free", "name": "OLMo 3.1 32B Think", "source": "openrouter", "provider": "AllenAI", "context_window": 128000, "supports_tool_calling": false, "supports_vision": false, "supports_structured_output": false, "description": "Open-source research model from AllenAI.", "pricing_note": "Free on OpenRouter", "best_for": "Research, open-source", "quality_tier": "mid", "status": "available", "first_seen_at": "2025-12-21T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "openrouter::agentica/deepcoder-14b-preview:free", "model_id": "agentica/deepcoder-14b-preview:free", "name": "DeepCoder 14B Preview", "source": "openrouter", "provider": "Agentica", "context_window": 128000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": false, "description": "Code generation focused 14B model.", "pricing_note": "Free on OpenRouter", "best_for": "Code generation", "quality_tier": "mid", "status": "available", "first_seen_at": "2026-01-30T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "provider::google-gemma-3n-e4b-it", "model_id": "gemma-3n-e4b-it", "name": "Gemma 3n E4B", "source": "provider-direct", "provider": "Google", "context_window": null, "supports_tool_calling": false, "supports_vision": true, "supports_structured_output": false, "description": "Efficient multimodal Gemma 3n model exposed through the Gemini API for text, image, video, and audio prototyping.", "pricing_note": "Free on Gemini Developer API", "best_for": "Efficient multimodal prototyping", "quality_tier": "mid", "status": "available", "first_seen_at": "2026-03-21T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "provider::google-gemini-robotics-er-1.5-preview", "model_id": "gemini-robotics-er-1.5-preview", "name": "Gemini Robotics-ER 1.5 Preview", "source": "provider-direct", "provider": "Google", "context_window": null, "supports_tool_calling": true, "supports_vision": true, "supports_structured_output": true, "description": "Gemini robotics model for spatial understanding, structured coordinates, and multi-step agentic task planning.", "pricing_note": "Free on Gemini Developer API", "best_for": "Robotics perception, spatial reasoning", "quality_tier": "high", "status": "available", "first_seen_at": "2026-03-21T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "provider::mistral-devstral-2", "model_id": "devstral-2512", "name": "Devstral 2", "source": "provider-direct", "provider": "Mistral", "context_window": null, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "Mistral frontier code-agent model for software engineering tasks, available on the Experiment plan.", "pricing_note": "Free on Mistral Experiment plan", "best_for": "Coding, software engineering tasks", "quality_tier": "high", "status": "available", "first_seen_at": "2026-03-21T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "kilo::minimax-m2.1", "model_id": "minimax/m2.1", "name": "MiniMax M2.1", "source": "kilo-gateway", "provider": "MiniMax", "context_window": 128000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "10B active params, strong coding via Kilo Code Gateway.", "pricing_note": "Free via Kilo Gateway", "best_for": "Coding, tool use", "quality_tier": "high", "status": "available", "first_seen_at": "2026-01-20T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "kilo::zai-glm-4.7", "model_id": "zai/glm-4.7", "name": "GLM 4.7", "source": "kilo-gateway", "provider": "Z.AI (Zhipu)", "context_window": 128000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "Latest GLM, agent-centric.", "pricing_note": "Free via Kilo Gateway", "best_for": "Agent workflows", "quality_tier": "high", "status": "available", "first_seen_at": "2025-12-21T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "kilo::moonshotai-kimi-k2.5", "model_id": "moonshotai/kimi-k2.5", "name": "Kimi K2.5", "source": "kilo-gateway", "provider": "MoonshotAI", "context_window": 128000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "Advanced tool use + reasoning.", "pricing_note": "Free via Kilo Gateway", "best_for": "Agentic workflows", "quality_tier": "high", "status": "available", "first_seen_at": "2026-02-27T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "kilo::giga-potato", "model_id": "stealth/giga-potato", "name": "Giga Potato", "source": "kilo-gateway", "provider": "Unknown (Stealth)", "context_window": null, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": false, "description": "Evaluation period stealth model.", "pricing_note": "Free via Kilo Gateway", "best_for": "Experimental", "quality_tier": "unknown", "status": "available", "first_seen_at": "2026-02-04T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "kilo::arcee-trinity", "model_id": "arcee-ai/trinity-large-preview", "name": "Arcee Trinity Large Preview", "source": "kilo-gateway", "provider": "Arcee AI", "context_window": 128000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "Also available on OpenRouter.", "pricing_note": "Free via Kilo Gateway", "best_for": "Tool use, agentic tasks", "quality_tier": "high", "status": "available", "first_seen_at": "2026-02-14T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "opencode::minimax-m2.5", "model_id": "minimax/m2.5-free", "name": "MiniMax M2.5 Free", "source": "opencode-zen", "provider": "MiniMax", "context_window": 128000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "Coding, agentic, Office file generation.", "pricing_note": "Free via OpenCode Zen", "best_for": "Coding, development", "quality_tier": "high", "status": "available", "first_seen_at": "2026-02-27T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "opencode::big-pickle", "model_id": "stealth/big-pickle", "name": "Big Pickle", "source": "opencode-zen", "provider": "Unknown (Stealth)", "context_window": null, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": false, "description": "Evaluation period stealth model.", "pricing_note": "Free via OpenCode Zen", "best_for": "Experimental", "quality_tier": "unknown", "status": "available", "first_seen_at": "2026-02-09T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "cline::minimax-m2.5", "model_id": "minimax/m2.5", "name": "MiniMax M2.5", "source": "cline", "provider": "MiniMax", "context_window": 128000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "80.2% SWE-Bench Verified, 100 TPS. Free in Cline CLI.", "pricing_note": "Free in Cline (limited time)", "best_for": "Coding, multi-agent workflows", "quality_tier": "high", "status": "available", "first_seen_at": "2026-02-27T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "provider::haimaker-minimax-m2.5", "model_id": "minimax/m2.5", "name": "MiniMax M2.5", "source": "provider-direct", "provider": "Haimaker.ai", "context_window": 128000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "Direct provider access via Haimaker.ai, currently marked free through April 1, 2026.", "pricing_note": "Free on Haimaker.ai until April 1, 2026", "best_for": "Coding, agent workflows", "quality_tier": "high", "status": "available", "first_seen_at": "2026-03-21T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" }, { "id": "cline::moonshotai-kimi-k2.5", "model_id": "moonshotai/kimi-k2.5", "name": "Kimi K2.5", "source": "cline", "provider": "MoonshotAI", "context_window": 128000, "supports_tool_calling": true, "supports_vision": false, "supports_structured_output": true, "description": "Free in Cline CLI 2.0.", "pricing_note": "Free in Cline CLI", "best_for": "Agentic coding, CLI workflows", "quality_tier": "high", "status": "available", "first_seen_at": "2026-02-27T04:14:32.000Z", "last_seen_at": "2026-03-21T04:14:32.000Z" } ], "providers": [ { "id": "google-gemini", "provider": "Google Gemini", "provider_slug": "google-gemini", "logo_url": null, "tier_name": "Free Tier", "free_credits": "Unlimited", "credit_expiry": "Never", "rate_limits": "15 RPM Flash, 5 RPM Pro, 1M ctx", "token_limit_monthly": "Generous", "credit_card_required": false, "models_available": "Gemini 3.1 Pro/Flash, Gemma 3/3n, Robotics-ER 1.5, embeddings", "data_training_policy": "May use for training on free tier", "signup_url": "https://aistudio.google.com/", "free_tier_url": "https://ai.google.dev/pricing", "docs_url": "https://ai.google.dev/docs", "notes": "Best free tier for most developers. No credit card, generous limits, plus free Gemma and robotics-preview access.", "last_verified_at": "2026-03-21T04:14:32.000Z" }, { "id": "xai-grok", "provider": "xAI (Grok)", "provider_slug": "xai", "logo_url": null, "tier_name": "Data Sharing Program", "free_credits": "$25 signup + $150/mo", "credit_expiry": "Never", "rate_limits": "Standard", "token_limit_monthly": "Generous", "credit_card_required": false, "models_available": "All Grok (4, 4.1 Fast, 3 Mini)", "data_training_policy": "Data sharing program uses interactions", "signup_url": "https://console.x.ai/", "free_tier_url": "https://docs.x.ai/docs/overview#api-credits", "docs_url": "https://docs.x.ai/", "notes": "Most generous free credits if you opt into data sharing.", "last_verified_at": "2026-03-19T04:14:32.000Z" }, { "id": "mistral", "provider": "Mistral", "provider_slug": "mistral", "logo_url": null, "tier_name": "Experiment Plan", "free_credits": "Unlimited", "credit_expiry": "Never", "rate_limits": "2 RPM, 500K TPM, 1B tokens/mo", "token_limit_monthly": "1B tokens", "credit_card_required": false, "models_available": "Mistral Large 3, Medium 3.1, Small 3.2, Devstral 2, Pixtral, Codestral", "data_training_policy": "EU data residency", "signup_url": "https://console.mistral.ai/", "free_tier_url": "https://mistral.ai/pricing", "docs_url": "https://docs.mistral.ai/", "notes": "EU-based. Experiment plan remains useful for testing current Mistral models with rate limits.", "last_verified_at": "2026-03-21T04:14:32.000Z" }, { "id": "openai", "provider": "OpenAI", "provider_slug": "openai", "logo_url": null, "tier_name": "Free Trial", "free_credits": "$5", "credit_expiry": "3 months", "rate_limits": "3 RPM on advanced models", "token_limit_monthly": "Credit-based", "credit_card_required": false, "models_available": "All (GPT-5.x, GPT-4o, o3, o4-mini)", "data_training_policy": "May use for training", "signup_url": "https://platform.openai.com/", "free_tier_url": "https://platform.openai.com/docs/guides/rate-limits/free-tier", "docs_url": "https://platform.openai.com/docs", "notes": "Credits expire after 3 months. Good for getting started.", "last_verified_at": "2026-03-20T04:14:32.000Z" }, { "id": "anthropic", "provider": "Anthropic", "provider_slug": "anthropic", "logo_url": null, "tier_name": "Free Credits", "free_credits": "~$5", "credit_expiry": "N/A", "rate_limits": "Standard", "token_limit_monthly": "Credit-based", "credit_card_required": false, "models_available": "All Claude (Opus 4.6, Sonnet 4.6, Haiku 4.5)", "data_training_policy": "Standard policy", "signup_url": "https://console.anthropic.com/", "free_tier_url": "https://docs.anthropic.com/en/docs/about-claude/pricing", "docs_url": "https://docs.anthropic.com/", "notes": "Phone verification required.", "last_verified_at": "2026-03-16T04:14:32.000Z" }, { "id": "together-ai", "provider": "Together AI", "provider_slug": "together", "logo_url": null, "tier_name": "Free Credits", "free_credits": "$5-$25 signup", "credit_expiry": "N/A", "rate_limits": "Standard", "token_limit_monthly": "Credit-based", "credit_card_required": false, "models_available": "200+ open-source (Llama, DeepSeek, Qwen, etc.)", "data_training_policy": "Standard policy", "signup_url": "https://api.together.xyz/", "free_tier_url": "https://www.together.ai/pricing", "docs_url": "https://docs.together.ai/", "notes": "Great selection of open-source models.", "last_verified_at": "2026-03-17T04:14:32.000Z" }, { "id": "groq", "provider": "Groq", "provider_slug": "groq", "logo_url": null, "tier_name": "Free Tier", "free_credits": "Free tier", "credit_expiry": "N/A", "rate_limits": "~30 RPM", "token_limit_monthly": "Rate-limited", "credit_card_required": false, "models_available": "Llama 3.3, Mixtral, Gemma 3", "data_training_policy": "Standard policy", "signup_url": "https://console.groq.com/", "free_tier_url": "https://groq.com/pricing", "docs_url": "https://console.groq.com/docs", "notes": "Fastest inference. Great for real-time apps.", "last_verified_at": "2026-03-19T04:14:32.000Z" }, { "id": "cohere", "provider": "Cohere", "provider_slug": "cohere", "logo_url": null, "tier_name": "Free Tier", "free_credits": "Free tier", "credit_expiry": "N/A", "rate_limits": "20 RPM", "token_limit_monthly": "Rate-limited", "credit_card_required": false, "models_available": "Command R+, Embed, Rerank", "data_training_policy": "Standard policy", "signup_url": "https://dashboard.cohere.com/", "free_tier_url": "https://cohere.com/pricing", "docs_url": "https://docs.cohere.com/", "notes": "Good for RAG and search use cases.", "last_verified_at": "2026-03-15T04:14:32.000Z" }, { "id": "hugging-face", "provider": "Hugging Face", "provider_slug": "huggingface", "logo_url": null, "tier_name": "Free Inference API", "free_credits": "Free tier", "credit_expiry": "N/A", "rate_limits": "Rate-limited, cold starts", "token_limit_monthly": "Rate-limited", "credit_card_required": false, "models_available": "100K+ open-source models", "data_training_policy": "Standard policy", "signup_url": "https://huggingface.co/", "free_tier_url": "https://huggingface.co/pricing", "docs_url": "https://huggingface.co/docs", "notes": "Largest model hub. Free Inference API with cold starts.", "last_verified_at": "2026-03-18T04:14:32.000Z" }, { "id": "replicate", "provider": "Replicate", "provider_slug": "replicate", "logo_url": null, "tier_name": "Signup Credits", "free_credits": "$5 signup credits", "credit_expiry": "14 days", "rate_limits": "Standard", "token_limit_monthly": "Credit-based", "credit_card_required": false, "models_available": "Llama, Stable Diffusion, Flux, etc.", "data_training_policy": "Standard policy", "signup_url": "https://replicate.com/", "free_tier_url": "https://replicate.com/pricing", "docs_url": "https://replicate.com/docs", "notes": "14-day expiry on credits.", "last_verified_at": "2026-03-16T04:14:32.000Z" }, { "id": "fireworks-ai", "provider": "Fireworks AI", "provider_slug": "fireworks", "logo_url": null, "tier_name": "Free Tier", "free_credits": "Free tier", "credit_expiry": "N/A", "rate_limits": "Rate-limited", "token_limit_monthly": "Rate-limited", "credit_card_required": false, "models_available": "Open-source models, fast inference", "data_training_policy": "Standard policy", "signup_url": "https://fireworks.ai/", "free_tier_url": "https://fireworks.ai/pricing", "docs_url": "https://docs.fireworks.ai/", "notes": "Fast inference on open-source models.", "last_verified_at": "2026-03-17T04:14:32.000Z" }, { "id": "sambanova", "provider": "SambaNova", "provider_slug": "sambanova", "logo_url": null, "tier_name": "Free Tier", "free_credits": "Free tier", "credit_expiry": "N/A", "rate_limits": "Rate-limited", "token_limit_monthly": "Rate-limited", "credit_card_required": false, "models_available": "Llama, open models on custom silicon", "data_training_policy": "Standard policy", "signup_url": "https://cloud.sambanova.ai/", "free_tier_url": "https://cloud.sambanova.ai/", "docs_url": "https://docs.sambanova.ai/", "notes": "Custom silicon for fast inference.", "last_verified_at": "2026-03-16T04:14:32.000Z" }, { "id": "cerebras", "provider": "Cerebras", "provider_slug": "cerebras", "logo_url": null, "tier_name": "Free Tier", "free_credits": "Free tier", "credit_expiry": "N/A", "rate_limits": "Rate-limited", "token_limit_monthly": "Rate-limited", "credit_card_required": false, "models_available": "Llama 3.3 70B, fast inference", "data_training_policy": "Standard policy", "signup_url": "https://cloud.cerebras.ai/", "free_tier_url": "https://cloud.cerebras.ai/", "docs_url": "https://docs.cerebras.ai/", "notes": "Fastest inference on Llama 3.3 70B.", "last_verified_at": "2026-03-17T04:14:32.000Z" }, { "id": "github-models", "provider": "GitHub Models", "provider_slug": "github-models", "logo_url": null, "tier_name": "Prototyping", "free_credits": "Free for prototyping", "credit_expiry": "N/A", "rate_limits": "Rate-limited", "token_limit_monthly": "Rate-limited", "credit_card_required": false, "models_available": "GPT-4o, Llama 3.1, Mistral, etc.", "data_training_policy": "Standard GitHub policy", "signup_url": "https://github.com/marketplace/models", "free_tier_url": "https://github.com/marketplace/models", "docs_url": "https://docs.github.com/en/github-models", "notes": "Requires GitHub account. Great for prototyping.", "last_verified_at": "2026-03-18T04:14:32.000Z" }, { "id": "cloudflare-ai", "provider": "Cloudflare Workers AI", "provider_slug": "cloudflare", "logo_url": null, "tier_name": "Workers AI Free", "free_credits": "10K tokens/day", "credit_expiry": "Daily reset", "rate_limits": "Daily reset", "token_limit_monthly": "~300K tokens/mo", "credit_card_required": false, "models_available": "Llama, Mistral, Phi on edge", "data_training_policy": "Standard Cloudflare policy", "signup_url": "https://ai.cloudflare.com/", "free_tier_url": "https://ai.cloudflare.com/", "docs_url": "https://developers.cloudflare.com/workers-ai/", "notes": "Edge inference, daily token reset.", "last_verified_at": "2026-03-16T04:14:32.000Z" }, { "id": "kilo-gateway", "provider": "Kilo Code Gateway", "provider_slug": "kilo-gateway", "logo_url": null, "tier_name": "Free Gateway", "free_credits": "Unlimited (curated models)", "credit_expiry": "Never", "rate_limits": "Fair use", "token_limit_monthly": "Unlimited", "credit_card_required": false, "models_available": "MiniMax M2.1, GLM 4.7, Kimi K2.5, Giga Potato, Arcee Trinity", "data_training_policy": "Varies by upstream provider", "signup_url": "https://kilocode.ai", "free_tier_url": "https://kilocode.ai", "docs_url": "https://docs.kilocode.ai", "notes": "Built into Kilo Code. Curated free models — no API key needed.", "last_verified_at": "2026-03-20T04:14:32.000Z" }, { "id": "opencode-zen", "provider": "OpenCode Zen", "provider_slug": "opencode-zen", "logo_url": null, "tier_name": "Free Gateway", "free_credits": "Unlimited (curated models)", "credit_expiry": "Never", "rate_limits": "Fair use", "token_limit_monthly": "Unlimited", "credit_card_required": false, "models_available": "MiniMax M2.5 Free, Big Pickle", "data_training_policy": "Varies by upstream provider", "signup_url": "https://opencode.ai", "free_tier_url": "https://opencode.ai", "docs_url": "https://docs.opencode.ai", "notes": "Built into OpenCode. Free model gateway — no API key needed.", "last_verified_at": "2026-03-19T04:14:32.000Z" }, { "id": "openrouter", "provider": "OpenRouter", "provider_slug": "openrouter", "logo_url": null, "tier_name": "Free Models", "free_credits": "Unlimited (free models)", "credit_expiry": "Never", "rate_limits": "~20 RPM, ~200 RPD", "token_limit_monthly": "Rate-limited", "credit_card_required": false, "models_available": "~29 free models (append :free to model ID)", "data_training_policy": "May train on data for free models", "signup_url": "https://openrouter.ai/", "free_tier_url": "https://openrouter.ai/models?q=free", "docs_url": "https://openrouter.ai/docs", "notes": "Largest selection of free models. Unified API. Free models marked with :free suffix.", "last_verified_at": "2026-03-21T04:14:32.000Z" }, { "id": "haimaker", "provider": "Haimaker.ai", "provider_slug": "haimaker", "logo_url": null, "tier_name": "Limited-Time Free Access", "free_credits": "MiniMax M2.5 free", "credit_expiry": "Until April 1, 2026", "rate_limits": "Undisclosed / fair use", "token_limit_monthly": "Undisclosed", "credit_card_required": false, "models_available": "MiniMax M2.5 (free until April 1, 2026)", "data_training_policy": "Standard policy", "signup_url": "https://app.haimaker.ai/sign-up", "free_tier_url": "https://app.haimaker.ai/sign-up", "docs_url": "https://haimaker.ai/", "notes": "Temporary free access to MiniMax M2.5. Availability appears promotional and may change after April 1, 2026.", "referral_url": "https://app.haimaker.ai/sign-up?ref=CKQHV8", "referral_note": "This sign-up link includes a referral code and may provide up to $50 in bonus credit.", "last_verified_at": "2026-03-21T04:14:32.000Z" } ], "agents": [ { "id": "github-copilot", "name": "GitHub Copilot Free", "logo_url": null, "description": "GitHub's AI pair programmer with free tier.", "free_plan_summary": "2,000 code completions + 50 premium requests per month.", "free_completions_monthly": "2,000", "free_premium_requests": "50/month", "included_models": "Claude 3.5 Sonnet, GPT-4o", "byok_supported": false, "open_source": false, "website_url": "https://github.com/features/copilot", "docs_url": "https://docs.github.com/copilot", "notes": "Free for verified students, teachers, and OSS maintainers. Includes Copilot CLI.", "last_verified_at": "2026-03-20T04:14:32.000Z" }, { "id": "gemini-code-assist", "name": "Gemini Code Assist", "logo_url": null, "description": "Google's AI coding assistant powered by Gemini.", "free_plan_summary": "180,000 completions/mo + unlimited chat.", "free_completions_monthly": "180,000", "free_premium_requests": "Unlimited chat", "included_models": "Gemini models", "byok_supported": false, "open_source": false, "website_url": "https://cloud.google.com/gemini/docs/code-assist", "docs_url": "https://cloud.google.com/gemini/docs", "notes": "Most generous free completions. Requires Google Cloud account.", "last_verified_at": "2026-03-19T04:14:32.000Z" }, { "id": "kilo-code", "name": "Kilo Code", "logo_url": null, "description": "Open-source AI coding agent with free gateway.", "free_plan_summary": "Free & open-source. Unlimited with BYOK + free Kilo Gateway models.", "free_completions_monthly": "Unlimited", "free_premium_requests": "Unlimited", "included_models": "Any via OpenRouter + free Kilo Gateway models", "byok_supported": true, "open_source": true, "website_url": "https://kilocode.ai", "docs_url": "https://docs.kilocode.ai", "notes": "Has its own free Gateway with curated models.", "last_verified_at": "2026-03-20T04:14:32.000Z" }, { "id": "cline", "name": "Cline", "logo_url": null, "description": "Open-source autonomous coding agent for VS Code and CLI.", "free_plan_summary": "Free & open-source. BYOK + free MiniMax M2.5 in CLI.", "free_completions_monthly": "Unlimited", "free_premium_requests": "Unlimited", "included_models": "Any via BYOK + free MiniMax M2.5 in CLI", "byok_supported": true, "open_source": true, "website_url": "https://cline.bot", "docs_url": "https://docs.cline.bot", "notes": "Also offers at-cost inference purchasing with no markup.", "last_verified_at": "2026-03-20T04:14:32.000Z" }, { "id": "roo-code", "name": "Roo Code", "logo_url": null, "description": "Open-source AI coding agent with Cloud Agents.", "free_plan_summary": "Free & open-source + Cloud Free tier with autonomous agents.", "free_completions_monthly": "Unlimited", "free_premium_requests": "Unlimited", "included_models": "Any via BYOK + Roo Provider at-cost models", "byok_supported": true, "open_source": true, "website_url": "https://roocode.com", "docs_url": "https://docs.roocode.com", "notes": "Cloud Agents run from GitHub, web, or Slack. No per-seat cost.", "last_verified_at": "2026-03-20T04:14:32.000Z" }, { "id": "opencode", "name": "OpenCode", "logo_url": null, "description": "Open-source coding agent with Zen gateway.", "free_plan_summary": "Free & open-source. Unlimited with BYOK + Zen free models.", "free_completions_monthly": "Unlimited", "free_premium_requests": "Unlimited", "included_models": "Any + Zen gateway free models", "byok_supported": true, "open_source": true, "website_url": "https://opencode.ai", "docs_url": "https://docs.opencode.ai", "notes": "Has its own Zen gateway with free models.", "last_verified_at": "2026-03-19T04:14:32.000Z" }, { "id": "claude-code", "name": "Claude Code", "logo_url": null, "description": "Anthropic's open-source terminal coding agent.", "free_plan_summary": "Free & open-source. Requires Anthropic API key (BYOK).", "free_completions_monthly": "Unlimited", "free_premium_requests": "Unlimited", "included_models": "Claude models only", "byok_supported": true, "open_source": true, "website_url": "https://docs.anthropic.com/en/docs/claude-code", "docs_url": "https://docs.anthropic.com/en/docs/claude-code", "notes": "Terminal-based. Claude models only via Anthropic key.", "last_verified_at": "2026-03-18T04:14:32.000Z" }, { "id": "aider", "name": "Aider", "logo_url": null, "description": "Open-source AI pair programming in your terminal.", "free_plan_summary": "Free & open-source. Unlimited with any OpenAI-compatible API.", "free_completions_monthly": "Unlimited", "free_premium_requests": "Unlimited", "included_models": "Any OpenAI-compatible API", "byok_supported": true, "open_source": true, "website_url": "https://aider.chat", "docs_url": "https://aider.chat/docs", "notes": "Terminal-based. Works with any OpenAI-compatible API.", "last_verified_at": "2026-03-18T04:14:32.000Z" }, { "id": "continue-dev", "name": "Continue.dev", "logo_url": null, "description": "Open-source AI code assistant for VS Code & JetBrains.", "free_plan_summary": "Free & open-source. Unlimited with BYOK.", "free_completions_monthly": "Unlimited", "free_premium_requests": "Unlimited", "included_models": "Any, VS Code + JetBrains", "byok_supported": true, "open_source": true, "website_url": "https://continue.dev", "docs_url": "https://docs.continue.dev", "notes": "Works with VS Code and JetBrains IDEs.", "last_verified_at": "2026-03-17T04:14:32.000Z" }, { "id": "cursor", "name": "Cursor", "logo_url": null, "description": "AI-first code editor with built-in AI features.", "free_plan_summary": "Hobby plan: limited Agent requests + limited Tab completions.", "free_completions_monthly": "2,000", "free_premium_requests": "50 premium", "included_models": "Claude, GPT-4, Gemini", "byok_supported": false, "open_source": false, "website_url": "https://cursor.com", "docs_url": "https://docs.cursor.com", "notes": "BYOK only on paid plans.", "last_verified_at": "2026-03-20T04:14:32.000Z" }, { "id": "windsurf", "name": "Windsurf", "logo_url": null, "description": "AI-powered IDE by Codeium.", "free_plan_summary": "Unlimited autocomplete + 25 prompt credits/mo.", "free_completions_monthly": "Unlimited", "free_premium_requests": "25 credits/mo", "included_models": "Gemini default, BYOK for Claude", "byok_supported": true, "open_source": false, "website_url": "https://windsurf.com", "docs_url": "https://docs.windsurf.com", "notes": "Unlimited Tab + inline edits on free plan.", "last_verified_at": "2026-03-20T04:14:32.000Z" }, { "id": "tabnine", "name": "Tabnine", "logo_url": null, "description": "AI code completion tool.", "free_plan_summary": "Free basic: unlimited basic completions.", "free_completions_monthly": "Unlimited basic", "free_premium_requests": "Limited", "included_models": "Own models", "byok_supported": false, "open_source": false, "website_url": "https://www.tabnine.com", "docs_url": "https://docs.tabnine.com", "notes": "Uses own models, no BYOK.", "last_verified_at": "2026-03-16T04:14:32.000Z" }, { "id": "amazon-codewhisperer", "name": "Amazon CodeWhisperer", "logo_url": null, "description": "AWS AI coding assistant.", "free_plan_summary": "Free: unlimited code suggestions.", "free_completions_monthly": "Unlimited", "free_premium_requests": "—", "included_models": "Amazon models", "byok_supported": false, "open_source": false, "website_url": "https://aws.amazon.com/codewhisperer/", "docs_url": "https://docs.aws.amazon.com/codewhisperer/", "notes": "Requires AWS account.", "last_verified_at": "2026-03-16T04:14:32.000Z" }, { "id": "v0", "name": "v0 (Vercel)", "logo_url": null, "description": "Vercel's AI-powered UI generation and full-stack coding agent.", "free_plan_summary": "Free tier with limited monthly credits.", "free_completions_monthly": "—", "free_premium_requests": "200 credits/month", "included_models": "Claude, GPT-4o", "byok_supported": false, "open_source": false, "website_url": "https://v0.dev", "docs_url": "https://v0.dev/docs", "notes": "Generates React/Next.js UI from prompts. Free tier includes chat and code generation.", "last_verified_at": "2026-03-20T04:14:32.000Z" } ], "changes": [ { "id": "16", "type": "new_model", "source": "openrouter", "subject": "Nemotron 3 Super", "title": "NVIDIA Nemotron 3 Super added to OpenRouter free tier", "details": "OpenRouter now exposes NVIDIA Nemotron 3 Super as a free model with a 1M-token context window for long-horizon reasoning and planning.", "detected_at": "2026-03-21T04:14:32.000Z" }, { "id": "15", "type": "new_model", "source": "provider-direct", "subject": "Gemini Robotics-ER 1.5 Preview", "title": "Google free tier now includes Gemini Robotics-ER 1.5 Preview", "details": "Gemini Developer API documentation now lists the robotics preview model for spatial reasoning, structured coordinates, and agentic robotics workflows.", "detected_at": "2026-03-21T04:14:32.000Z" }, { "id": "14", "type": "new_model", "source": "provider-direct", "subject": "Gemma 3n E4B", "title": "Google Gemma 3n E4B available on Gemini Developer API free tier", "details": "Google exposes Gemma 3n through the Gemini API, giving the directory a direct-provider multimodal Gemma option beyond the OpenRouter entries.", "detected_at": "2026-03-21T04:14:32.000Z" }, { "id": "13", "type": "new_model", "source": "provider-direct", "subject": "Devstral 2", "title": "Mistral Experiment plan includes Devstral 2", "details": "Mistral documents Devstral 2 as part of the current model catalog while keeping free API usage available on the Experiment plan.", "detected_at": "2026-03-21T04:14:32.000Z" }, { "id": "12", "type": "new_provider_tier", "source": "provider-direct", "subject": "Haimaker.ai", "title": "Haimaker.ai added with MiniMax M2.5 free through April 1, 2026", "details": "Haimaker.ai now appears in the directory with direct MiniMax M2.5 access marked free until April 1, 2026, plus a transparent referral disclosure on signup.", "detected_at": "2026-03-21T04:14:32.000Z" }, { "id": "1", "type": "new_model", "source": "openrouter", "subject": "Qwen3 Coder 480B", "title": "Qwen3 Coder 480B A35B added to OpenRouter free tier", "details": "Qwen's massive 480B coding model (A35B MoE) is now free on OpenRouter with 262K context and tool calling support.", "detected_at": "2026-03-07T04:14:32.000Z" }, { "id": "2", "type": "new_model", "source": "openrouter", "subject": "StepFun Step 3.5 Flash", "title": "StepFun Step 3.5 Flash joins OpenRouter free models", "details": "StepFun's Step 3.5 Flash model is now available for free on OpenRouter with 256K context and tool calling.", "detected_at": "2026-03-03T04:14:32.000Z" }, { "id": "3", "type": "new_model", "source": "openrouter", "subject": "NVIDIA Nemotron 3 Nano 30B", "title": "NVIDIA Nemotron 3 Nano 30B A3B free on OpenRouter", "details": "NVIDIA's Nemotron 3 Nano 30B with A3B architecture added to OpenRouter free tier. 256K context with tool calling.", "detected_at": "2026-02-28T04:14:32.000Z" }, { "id": "4", "type": "agent_update", "source": "cline", "subject": "Cline CLI 2.0", "title": "Cline CLI 2.0 launches with free MiniMax M2.5 and Kimi K2.5", "details": "Cline released CLI 2.0 with free access to MiniMax M2.5 (100 TPS) and Kimi K2.5 — no API key required.", "detected_at": "2026-02-27T04:14:32.000Z" }, { "id": "5", "type": "new_model", "source": "openrouter", "subject": "Qwen3-Next-80B-A3B", "title": "Qwen3-Next-80B-A3B now free on OpenRouter", "details": "Qwen's next-gen 80B model with A3B MoE architecture joins the free tier. 262K context with tool calling.", "detected_at": "2026-02-24T04:14:32.000Z" }, { "id": "6", "type": "new_model", "source": "openrouter", "subject": "Arcee Trinity", "title": "Arcee AI Trinity Large + Mini added to OpenRouter free tier", "details": "Arcee AI's Trinity models join OpenRouter free. Large (400B/13B MoE) and Mini (26B/3B MoE) with tool calling.", "detected_at": "2026-02-21T04:14:32.000Z" }, { "id": "7", "type": "agent_update", "source": "github", "subject": "GitHub Copilot Free", "title": "GitHub Copilot Free now includes GPT-5 mini and Haiku 4.5", "details": "GitHub updated Copilot Free to include access to GPT-5 mini and Claude Haiku 4.5. Still 2,000 completions and 50 premium requests/month.", "detected_at": "2026-02-19T04:14:32.000Z" }, { "id": "8", "type": "new_model", "source": "openrouter", "subject": "Free Models Router", "title": "OpenRouter launches Free Models Router", "details": "OpenRouter introduced openrouter/free — a meta-router that auto-selects from available free models. 200K context, supports vision and tool calling.", "detected_at": "2026-02-15T04:14:32.000Z" }, { "id": "9", "type": "tier_changed", "source": "xai", "subject": "xAI (Grok)", "title": "xAI offers $150/month in free API credits via data sharing", "details": "xAI's data sharing program now provides $150/month in free API credits plus $25 signup bonus. Covers all Grok models.", "detected_at": "2026-02-14T04:14:32.000Z" }, { "id": "10", "type": "agent_update", "source": "roo-code", "subject": "Roo Code Cloud", "title": "Roo Code launches free Cloud tier with autonomous agents", "details": "Roo Code introduced a free Cloud tier with autonomous Cloud Agents from GitHub, web, or Slack.", "detected_at": "2026-02-04T04:14:32.000Z" }, { "id": "11", "type": "agent_update", "source": "cursor", "subject": "Cursor", "title": "Cursor replaces specific free limits with \"limited\" usage model", "details": "Cursor moved from specific request counts to \"limited Agent requests\" and \"limited Tab completions\" on free plan.", "detected_at": "2026-01-25T04:14:32.000Z" } ] }