{ "version": "1.0.0", "generated": "2026-05-24T01:00:00Z", "source": "ollama-cloud-models-v2026-05-24", "total_agents": 34, "total_models_tracked": 13, "providers": ["ollama-cloud"], "models": [ { "id": "deepseek-v4-pro-max", "name": "DeepSeek V4-Pro Max", "organization": "DeepSeek", "parameters": "1.6T/49B active MoE", "context_window": "1M", "swe_bench": 80.6, "if_score": 89, "categories": ["coding", "agent", "reasoning"], "provider": "ollama-cloud", "updated": "2026-05-03", "pulls": "71.6K" }, { "id": "deepseek-v4-flash", "name": "DeepSeek V4-Flash", "organization": "DeepSeek", "parameters": "284B/13B active MoE", "context_window": "1M", "swe_bench": 79, "if_score": 86, "categories": ["coding", "efficient", "agent"], "provider": "ollama-cloud", "updated": "2026-05-03", "pulls": "84.4K" }, { "id": "kimi-k2.6", "name": "Kimi K2.6", "organization": "Moonshot AI", "parameters": "1T/32B active MoE", "context_window": "256K→1M", "swe_bench": 80.2, "if_score": 91, "categories": ["coding", "agent", "multimodal", "vision"], "provider": "ollama-cloud", "updated": "2026-04-24", "pulls": "259.7K" }, { "id": "kimi-k2.5", "name": "Kimi K2.5", "organization": "Moonshot AI", "parameters": "1T/32B active MoE", "context_window": "256K", "swe_bench": 78, "if_score": 90, "categories": ["coding", "agent", "multimodal", "vision"], "provider": "ollama-cloud", "updated": "2026-02-24", "pulls": "293.2K" }, { "id": "qwen3-coder-480b", "name": "Qwen3-Coder 480B", "organization": "Qwen", "parameters": "480B/35B active", "context_window": "256K→1M", "swe_bench": 66.5, "if_score": 88, "categories": ["coding", "agent"], "provider": "ollama-cloud", "updated": "2026-02-24", "pulls": "N/A (legacy track)" }, { "id": "qwen3.5-122b", "name": "Qwen 3.5 122B", "organization": "Qwen", "parameters": "122B/10B active", "context_window": "128K", "swe_bench": null, "if_score": 92, "categories": ["reasoning", "efficient", "vision", "tools"], "provider": "ollama-cloud", "updated": "2026-05-22", "pulls": "12.4M" }, { "id": "gemma4-27b", "name": "Gemma 4 (27B)", "organization": "Google", "parameters": "27B", "context_window": "128K", "swe_bench": null, "if_score": 85, "categories": ["coding", "agent", "reasoning", "vision", "audio"], "provider": "ollama-cloud", "updated": "2026-05-22", "pulls": "10.1M", "note": "Updated 2 days ago. Frontier-level performance at each size." }, { "id": "minimax-m2.5", "name": "MiniMax M2.5", "organization": "MiniMax", "parameters": "MoE undisclosed", "context_window": "128K", "swe_bench": 80.2, "if_score": 82, "categories": ["coding", "agent"], "provider": "ollama-cloud", "updated": "2026-02-24", "pulls": "2.2M" }, { "id": "minimax-m2.7", "name": "MiniMax M2.7", "organization": "MiniMax", "parameters": "~10B active", "context_window": "128K", "swe_bench": 78, "if_score": 80, "categories": ["coding", "agent", "efficient"], "provider": "ollama-cloud", "updated": "2026-03-24", "pulls": "2.2M" }, { "id": "glm-5.1", "name": "GLM-5.1", "organization": "Z.ai", "parameters": "744B/40B active", "context_window": "128K", "swe_bench": null, "if_score": 90, "categories": ["reasoning", "agent"], "provider": "ollama-cloud", "updated": "2026-04-24", "pulls": "2.2M", "note": "Next-gen flagship. SWE-Bench Pro SOTA." }, { "id": "glm-5", "name": "GLM-5", "organization": "Z.ai", "parameters": "744B/40B active", "context_window": "128K", "swe_bench": null, "if_score": 90, "categories": ["reasoning", "agent"], "provider": "ollama-cloud", "updated": "2026-02-24", "pulls": "2.3M" }, { "id": "nemotron-3-super", "name": "Nemotron 3 Super", "organization": "NVIDIA", "parameters": "120B/12B active", "context_window": "1M", "swe_bench": 60.5, "if_score": 78, "categories": ["agent", "reasoning", "efficient"], "provider": "ollama-cloud", "updated": "2026-03-24", "pulls": "2.4M" }, { "id": "nemotron-3-nano", "name": "Nemotron 3 Nano", "organization": "NVIDIA", "parameters": "30B/4B", "context_window": "128K", "swe_bench": null, "if_score": 68, "categories": ["agent", "efficient"], "provider": "ollama-cloud", "updated": "2026-03-24", "pulls": "453K" }, { "id": "devstral-2", "name": "Devstral 2", "organization": "Mistral / Devstral", "parameters": "123B", "context_window": "128K", "swe_bench": null, "if_score": 80, "categories": ["coding", "agent"], "provider": "ollama-cloud", "updated": "2026-02-24", "pulls": "223.2K" }, { "id": "devstral-small-2", "name": "Devstral Small 2", "organization": "Mistral / Devstral", "parameters": "24B", "context_window": "128K", "swe_bench": null, "if_score": 75, "categories": ["coding", "agent"], "provider": "ollama-cloud", "updated": "2026-02-24", "pulls": "838.8K" } ], "if_scores": { "deepseek-v4-pro-max": 89, "deepseek-v4-flash": 86, "kimi-k2.6": 91, "kimi-k2.5": 90, "qwen3-coder-480b": 88, "qwen3.5-122b": 92, "gemma4-27b": 85, "minimax-m2.5": 82, "minimax-m2.7": 80, "glm-5.1": 90, "glm-5": 90, "nemotron-3-super": 78, "nemotron-3-nano": 68, "devstral-2": 80, "devstral-small-2": 75 }, "agent_model_scores": [ { "agent": "lead-developer", "current_model_index": 0, "scores": { "qwen3-coder-480b": 92, "deepseek-v4-pro-max": 88, "deepseek-v4-flash": 85, "kimi-k2.6": 90, "kimi-k2.5": 88, "qwen3.5-122b": 86, "gemma4-27b": 83, "minimax-m2.5": 86, "minimax-m2.7": 82, "glm-5.1": 68, "nemotron-3-super": 70, "devstral-2": 84, "devstral-small-2": 78 } }, { "agent": "frontend-developer", "scores": { "qwen3-coder-480b": 86, "deepseek-v4-pro-max": 82, "deepseek-v4-flash": 80, "kimi-k2.6": 86, "kimi-k2.5": 84, "qwen3.5-122b": 84, "gemma4-27b": 85, "minimax-m2.5": 92, "minimax-m2.7": 88, "glm-5.1": 56, "nemotron-3-super": 62, "devstral-2": 80, "devstral-small-2": 74 } }, { "agent": "backend-developer", "scores": { "qwen3-coder-480b": 91, "deepseek-v4-pro-max": 86, "kimi-k2.6": 90, "qwen3.5-122b": 85, "gemma4-27b": 84, "minimax-m2.5": 84, "minimax-m2.7": 80, "glm-5.1": 63, "nemotron-3-super": 68, "devstral-2": 82, "devstral-small-2": 76 } }, { "agent": "go-developer", "scores": { "qwen3-coder-480b": 85, "deepseek-v4-pro-max": 88, "deepseek-v4-flash": 84, "kimi-k2.6": 86, "qwen3.5-122b": 80, "gemma4-27b": 80, "minimax-m2.5": 78, "minimax-m2.7": 74, "glm-5.1": 58, "nemotron-3-super": 66, "devstral-2": 82, "devstral-small-2": 74 } }, { "agent": "python-developer", "scores": { "qwen3-coder-480b": 90, "deepseek-v4-pro-max": 78, "kimi-k2.6": 88, "qwen3.5-122b": 86, "gemma4-27b": 82, "minimax-m2.5": 82, "minimax-m2.7": 78, "glm-5.1": 60, "nemotron-3-super": 66, "devstral-2": 86, "devstral-small-2": 80 } }, { "agent": "php-developer", "scores": { "qwen3-coder-480b": 87, "deepseek-v4-pro-max": 74, "kimi-k2.6": 86, "qwen3.5-122b": 84, "gemma4-27b": 82, "minimax-m2.5": 76, "minimax-m2.7": 72, "glm-5.1": 56, "nemotron-3-super": 64, "devstral-2": 80, "devstral-small-2": 74 } }, { "agent": "devops-engineer", "scores": { "qwen3-coder-480b": 66, "deepseek-v4-pro-max": 80, "kimi-k2.6": 88, "qwen3.5-122b": 75, "gemma4-27b": 78, "minimax-m2.5": 53, "minimax-m2.7": 48, "glm-5.1": 75, "nemotron-3-super": 78, "devstral-2": 72, "devstral-small-2": 68 } }, { "agent": "sdet-engineer", "scores": { "qwen3-coder-480b": 88, "deepseek-v4-pro-max": 84, "kimi-k2.6": 87, "qwen3.5-122b": 86, "gemma4-27b": 82, "minimax-m2.5": 84, "minimax-m2.7": 80, "glm-5.1": 63, "nemotron-3-super": 70, "devstral-2": 86, "devstral-small-2": 80 } }, { "agent": "code-skeptic", "scores": { "qwen3-coder-480b": 82, "deepseek-v4-pro-max": 82, "kimi-k2.6": 82, "qwen3.5-122b": 80, "gemma4-27b": 80, "minimax-m2.5": 85, "minimax-m2.7": 80, "glm-5.1": 72, "nemotron-3-super": 73, "devstral-2": 82, "devstral-small-2": 76 } }, { "agent": "security-auditor", "scores": { "qwen3-coder-480b": 76, "deepseek-v4-pro-max": 80, "kimi-k2.6": 80, "qwen3.5-122b": 78, "gemma4-27b": 78, "minimax-m2.5": 74, "minimax-m2.7": 68, "glm-5.1": 68, "nemotron-3-super": 76, "devstral-2": 78, "devstral-small-2": 72 } }, { "agent": "performance-engineer", "scores": { "qwen3-coder-480b": 78, "deepseek-v4-pro-max": 84, "kimi-k2.6": 82, "qwen3.5-122b": 76, "gemma4-27b": 76, "minimax-m2.5": 75, "minimax-m2.7": 70, "glm-5.1": 74, "nemotron-3-super": 78, "devstral-2": 80, "devstral-small-2": 74 } }, { "agent": "the-fixer", "scores": { "qwen3-coder-480b": 89, "deepseek-v4-pro-max": 88, "kimi-k2.6": 90, "qwen3.5-122b": 86, "gemma4-27b": 82, "minimax-m2.5": 88, "minimax-m2.7": 84, "glm-5.1": 64, "nemotron-3-super": 71, "devstral-2": 86, "devstral-small-2": 82 } }, { "agent": "browser-automation", "scores": { "qwen3-coder-480b": 87, "deepseek-v4-pro-max": 82, "kimi-k2.6": 86, "qwen3.5-122b": 82, "gemma4-27b": 84, "minimax-m2.5": 72, "minimax-m2.7": 68, "glm-5.1": 53, "nemotron-3-super": 61, "devstral-2": 80, "devstral-small-2": 74 } }, { "agent": "visual-tester", "scores": { "qwen3-coder-480b": 82, "deepseek-v4-pro-max": 76, "kimi-k2.6": 78, "qwen3.5-122b": 76, "gemma4-27b": 78, "minimax-m2.5": 68, "minimax-m2.7": 64, "glm-5.1": 48, "nemotron-3-super": 55, "devstral-2": 74, "devstral-small-2": 68 } }, { "agent": "system-analyst", "scores": { "qwen3-coder-480b": 70, "deepseek-v4-pro-max": 88, "kimi-k2.6": 86, "qwen3.5-122b": 82, "gemma4-27b": 82, "minimax-m2.5": 66, "minimax-m2.7": 63, "glm-5.1": 82, "nemotron-3-super": 74, "devstral-2": 80, "devstral-small-2": 74 } }, { "agent": "capability-analyst", "scores": { "qwen3-coder-480b": 72, "deepseek-v4-pro-max": 82, "kimi-k2.6": 82, "qwen3.5-122b": 80, "gemma4-27b": 80, "minimax-m2.5": 68, "minimax-m2.7": 66, "glm-5.1": 78, "nemotron-3-super": 76, "devstral-2": 78, "devstral-small-2": 72 } }, { "agent": "orchestrator", "scores": { "qwen3-coder-480b": 74, "deepseek-v4-pro-max": 86, "kimi-k2.6": 92, "qwen3.5-122b": 84, "gemma4-27b": 82, "minimax-m2.5": 70, "minimax-m2.7": 68, "glm-5.1": 82, "nemotron-3-super": 80, "devstral-2": 80, "devstral-small-2": 74 } }, { "agent": "release-manager", "scores": { "qwen3-coder-480b": 72, "deepseek-v4-pro-max": 78, "kimi-k2.6": 78, "qwen3.5-122b": 76, "gemma4-27b": 76, "minimax-m2.5": 66, "minimax-m2.7": 64, "glm-5.1": 76, "nemotron-3-super": 74, "devstral-2": 76, "devstral-small-2": 70 } }, { "agent": "evaluator", "scores": { "qwen3-coder-480b": 70, "deepseek-v4-pro-max": 84, "kimi-k2.6": 84, "qwen3.5-122b": 82, "gemma4-27b": 80, "minimax-m2.5": 73, "minimax-m2.7": 70, "glm-5.1": 78, "nemotron-3-super": 78, "devstral-2": 80, "devstral-small-2": 74 } }, { "agent": "prompt-optimizer", "scores": { "qwen3-coder-480b": 76, "deepseek-v4-pro-max": 80, "kimi-k2.6": 82, "qwen3.5-122b": 82, "gemma4-27b": 80, "minimax-m2.5": 74, "minimax-m2.7": 72, "glm-5.1": 75, "nemotron-3-super": 76, "devstral-2": 80, "devstral-small-2": 74 } }, { "agent": "product-owner", "scores": { "qwen3-coder-480b": 60, "deepseek-v4-pro-max": 76, "kimi-k2.6": 76, "qwen3.5-122b": 76, "gemma4-27b": 76, "minimax-m2.5": 56, "minimax-m2.7": 54, "glm-5.1": 78, "nemotron-3-super": 74, "devstral-2": 76, "devstral-small-2": 70 } }, { "agent": "pipeline-judge", "scores": { "qwen3-coder-480b": 64, "deepseek-v4-pro-max": 82, "kimi-k2.6": 84, "qwen3.5-122b": 82, "gemma4-27b": 80, "minimax-m2.5": 68, "minimax-m2.7": 65, "glm-5.1": 76, "nemotron-3-super": 78, "devstral-2": 78, "devstral-small-2": 72 } }, { "agent": "workflow-architect", "scores": { "qwen3-coder-480b": 68, "deepseek-v4-pro-max": 80, "kimi-k2.6": 82, "qwen3.5-122b": 80, "gemma4-27b": 80, "minimax-m2.5": 62, "minimax-m2.7": 60, "glm-5.1": 76, "nemotron-3-super": 76, "devstral-2": 78, "devstral-small-2": 72 } }, { "agent": "markdown-validator", "scores": { "qwen3-coder-480b": 43, "deepseek-v4-pro-max": 68, "kimi-k2.6": 56, "qwen3.5-122b": 56, "gemma4-27b": 60, "minimax-m2.5": 38, "minimax-m2.7": 36, "glm-5.1": 55, "nemotron-3-super": 52, "nemotron-3-nano": 70, "devstral-2": 65, "devstral-small-2": 62 } }, { "agent": "agent-architect", "scores": { "qwen3-coder-480b": 78, "deepseek-v4-pro-max": 82, "kimi-k2.6": 86, "qwen3.5-122b": 80, "gemma4-27b": 82, "minimax-m2.5": 72, "minimax-m2.7": 70, "glm-5.1": 76, "nemotron-3-super": 78, "devstral-2": 80, "devstral-small-2": 74 } }, { "agent": "planner", "scores": { "qwen3-coder-480b": 72, "deepseek-v4-pro-max": 88, "kimi-k2.6": 86, "qwen3.5-122b": 86, "gemma4-27b": 84, "minimax-m2.5": 68, "minimax-m2.7": 66, "glm-5.1": 78, "nemotron-3-super": 80, "devstral-2": 84, "devstral-small-2": 78 } }, { "agent": "reflector", "scores": { "qwen3-coder-480b": 68, "deepseek-v4-pro-max": 84, "kimi-k2.6": 80, "qwen3.5-122b": 80, "gemma4-27b": 80, "minimax-m2.5": 66, "minimax-m2.7": 64, "glm-5.1": 76, "nemotron-3-super": 78, "devstral-2": 82, "devstral-small-2": 76 } }, { "agent": "memory-manager", "scores": { "qwen3-coder-480b": 63, "deepseek-v4-pro-max": 86, "kimi-k2.6": 84, "qwen3.5-122b": 85, "gemma4-27b": 82, "minimax-m2.5": 58, "minimax-m2.7": 56, "glm-5.1": 72, "nemotron-3-super": 86, "devstral-2": 78, "devstral-small-2": 72 } }, { "agent": "architect-indexer", "scores": { "qwen3-coder-480b": 70, "deepseek-v4-pro-max": 78, "kimi-k2.6": 84, "qwen3.5-122b": 80, "gemma4-27b": 80, "minimax-m2.5": 64, "minimax-m2.7": 62, "glm-5.1": 80, "nemotron-3-super": 74, "devstral-2": 78, "devstral-small-2": 72 } }, { "agent": "flutter-developer", "scores": { "qwen3-coder-480b": 86, "deepseek-v4-pro-max": 78, "kimi-k2.6": 84, "qwen3.5-122b": 84, "gemma4-27b": 84, "minimax-m2.5": 70, "minimax-m2.7": 66, "glm-5.1": 53, "nemotron-3-super": 60, "devstral-2": 78, "devstral-small-2": 74 } } ], "agent_current_config": [ { "agent": "lead-developer", "model": "ollama-cloud/qwen3-coder:480b", "fit_score": 92, "status": "optimal" }, { "agent": "frontend-developer", "model": "ollama-cloud/minimax-m2.5", "fit_score": 92, "status": "optimal" }, { "agent": "backend-developer", "model": "ollama-cloud/qwen3-coder:480b", "fit_score": 91, "status": "optimal" }, { "agent": "go-developer", "model": "ollama-cloud/deepseek-v4-pro-max", "fit_score": 88, "status": "optimal" }, { "agent": "python-developer", "model": "ollama-cloud/qwen3-coder:480b", "fit_score": 90, "status": "optimal" }, { "agent": "php-developer", "model": "ollama-cloud/qwen3-coder:480b", "fit_score": 87, "status": "optimal" }, { "agent": "flutter-developer", "model": "ollama-cloud/qwen3-coder:480b", "fit_score": 86, "status": "optimal" }, { "agent": "devops-engineer", "model": "ollama-cloud/kimi-k2.6", "fit_score": 88, "status": "optimal" }, { "agent": "sdet-engineer", "model": "ollama-cloud/qwen3-coder:480b", "fit_score": 88, "status": "optimal" }, { "agent": "code-skeptic", "model": "ollama-cloud/minimax-m2.5", "fit_score": 85, "status": "optimal" }, { "agent": "security-auditor", "model": "ollama-cloud/deepseek-v4-pro-max", "fit_score": 80, "status": "good" }, { "agent": "performance-engineer", "model": "ollama-cloud/deepseek-v4-pro-max", "fit_score": 84, "status": "optimal" }, { "agent": "the-fixer", "model": "ollama-cloud/kimi-k2.6", "fit_score": 90, "status": "optimal" }, { "agent": "browser-automation", "model": "ollama-cloud/qwen3-coder:480b", "fit_score": 87, "status": "optimal" }, { "agent": "visual-tester", "model": "ollama-cloud/qwen3-coder:480b", "fit_score": 82, "status": "good" }, { "agent": "system-analyst", "model": "ollama-cloud/glm-5.1", "fit_score": 82, "status": "good" }, { "agent": "capability-analyst", "model": "ollama-cloud/glm-5.1", "fit_score": 78, "status": "good" }, { "agent": "orchestrator", "model": "ollama-cloud/kimi-k2.6", "fit_score": 92, "status": "optimal" }, { "agent": "release-manager", "model": "ollama-cloud/glm-5.1", "fit_score": 76, "status": "good" }, { "agent": "evaluator", "model": "ollama-cloud/glm-5.1", "fit_score": 78, "status": "good" }, { "agent": "prompt-optimizer", "model": "ollama-cloud/qwen3.5", "fit_score": 82, "status": "recommended" }, { "agent": "product-owner", "model": "ollama-cloud/glm-5.1", "fit_score": 78, "status": "good" }, { "agent": "pipeline-judge", "model": "ollama-cloud/glm-5.1", "fit_score": 76, "status": "good" }, { "agent": "workflow-architect", "model": "ollama-cloud/glm-5.1", "fit_score": 76, "status": "good" }, { "agent": "markdown-validator", "model": "ollama-cloud/deepseek-v4-pro-max", "fit_score": 68, "status": "poor" }, { "agent": "agent-architect", "model": "ollama-cloud/kimi-k2.6", "fit_score": 86, "status": "optimal" }, { "agent": "planner", "model": "ollama-cloud/deepseek-v4-pro-max", "fit_score": 88, "status": "optimal" }, { "agent": "reflector", "model": "ollama-cloud/deepseek-v4-pro-max", "fit_score": 84, "status": "optimal" }, { "agent": "memory-manager", "model": "ollama-cloud/qwen3.5", "fit_score": 85, "status": "recommended" }, { "agent": "architect-indexer", "model": "ollama-cloud/glm-5.1", "fit_score": 80, "status": "good" } ], "recommendations": [ { "agent": "prompt-optimizer", "from_model": "ollama-cloud/qwen3.6-plus (openrouter)", "to_model": "ollama-cloud/qwen3.5", "reason": "Migrated to Ollama Cloud. IF 92, vision+tools+thinking. Same quality, no rate limits.", "impact": "high", "applied": false }, { "agent": "memory-manager", "from_model": "ollama-cloud/qwen3.6-plus (openrouter)", "to_model": "ollama-cloud/qwen3.5", "reason": "Migrated to Ollama Cloud. 1M context via qwen3.5? Actually qwen3.5 has 128K, not 1M. Alternative: kimi-k2.6 (256K) or deepseek-v4 (1M). But matrix shows qwen3.5=85 vs kimi-k2.6=84 vs deepseek=86.", "impact": "high", "applied": false }, { "agent": "markdown-validator", "from_model": "ollama-cloud/deepseek-v4-pro-max", "to_model": "ollama-cloud/nemotron-3-nano", "reason": "Markdown validator scores are lowest (68 max). Nemotron-3-Nano IF=68 but is tiny (4B/30B), extremely cheap. For lightweight validation tasks, nano is sufficient.", "impact": "medium", "applied": false }, { "agent": "markdown-validator", "from_model": "ollama-cloud/deepseek-v4-pro-max", "to_model": "ollama-cloud/gemma4-27b", "reason": "Gemma 4 is newest (2 days), frontier at each size. Scores 60 for validator — better than nano 70? Actually wait: gemma4=60, nano=70. Nano is better for this role. But gemma4 is newer and more general.", "impact": "low", "applied": false }, { "agent": "system-analyst", "from_model": "ollama-cloud/glm-5.1", "to_model": "ollama-cloud/deepseek-v4-pro-max", "reason": "Matrix: deepseek-v4-pro-max=88 vs glm-5.1=82. +6% quality, 1M context for architecture docs. GLM-5.1 still strong for standardization.", "impact": "medium", "applied": false }, { "agent": "evaluator", "from_model": "ollama-cloud/glm-5.1", "to_model": "ollama-cloud/kimi-k2.6", "reason": "Matrix: kimi-k2.6=84 vs glm-5.1=78. +6%. IF=91 for scoring accuracy. High reasoning needed.", "impact": "medium", "applied": false }, { "agent": "evaluator", "from_model": "ollama-cloud/glm-5.1", "to_model": "ollama-cloud/deepseek-v4-pro-max", "reason": "Alternative to kimi-k2.6. deepseek-v4-pro-max=84 (same as kimi), but 1M context. Could be better for large evaluation tasks.", "impact": "medium", "applied": false }, { "agent": "security-auditor", "from_model": "ollama-cloud/deepseek-v4-pro-max", "to_model": "ollama-cloud/kimi-k2.6", "reason": "Matrix: both 80. But kimi-k2.6 has multimodal (vision) which could help with screenshot-based security analysis. Tie.", "impact": "low", "applied": false }, { "agent": "gemma4-trial", "from_model": "none", "to_model": "ollama-cloud/gemma4-27b", "reason": "Gemma 4 is brand new (2 days), 10.1M pulls, frontier at each size, vision+audio+thinking. Could be game-changer for frontend-dev, browser-automation, visual-tester.", "impact": "high", "applied": false, "note": "Requires A/B test on frontend task." }, { "agent": "qwen3.5-trial", "from_model": "none", "to_model": "ollama-cloud/qwen3.5-122b", "reason": "Qwen 3.5 updated 2 days ago, 12.4M pulls, IF=92 (highest!), multimodal. Could replace GLM-5.1 for reasoning tasks and qwen3-coder for some coding tasks.", "impact": "high", "applied": false, "note": "Requires A/B test on planner/evaluator tasks." } ], "new_models_to_consider": [ { "id": "gemma4-27b", "priority": "critical", "rationale": "Updated 2 days ago. 10.1M pulls. Frontier-level at each size. Vision + audio + thinking + tools + cloud. Potentially replaces qwen3-coder for some tasks." }, { "id": "qwen3.5-122b", "priority": "critical", "rationale": "Updated 2 days ago. 12.4M pulls. IF=92 highest among tracked. Multimodal. Could replace glm-5.1 for reasoning and compete with qwen3-coder for coding." }, { "id": "deepseek-v4-flash", "priority": "medium", "rationale": "Same family as pro-max but much faster (13B active vs 49B). Good for low-latency agents: code-skeptic, browser-automation." }, { "id": "devstral-2", "priority": "medium", "rationale": "123B model for tool use and codebase exploration. Could be strong for lead-developer on large projects." } ] }