{"slug": "gpt-5-4-mini", "name": "GPT-5.4 mini", "category": "ai", "category_name": "AI models and assistants", "description": "A smaller, lower-cost OpenAI language model in the GPT-5.4 family.", "recommenders": 5, "commenters": 6, "warners": 1, "score": 3.694, "rising": 1.02, "first": "2026-04-04", "latest": "2026-09-14", "why": "HN commenters like it as a fast, token-efficient model for implementing coding features, and some find it good at code review.", "why_is_quote": false, "has_verdict": true, "sentiment": {"recommend": 5, "neutral": 0, "compare": 0, "warn": 1}, "spark": [0, 0, 0, 0, 0, 1, 0, 2, 0, 0, 2, 0], "listed": true, "loved": ["Coding implementation", "Speed"], "caveat": "Needs guardrails", "short_quote": "I switched to GPT-5.4-mini, which uses far fewer tokens, and it\u2019s often just as good as Sonnet.", "icon": false, "domain": null, "github_repo": null, "official_url": "", "pricing": "paid", "aliases": ["GPT-5.4 mini", "GPT-5.4-mini", "GPT 5.4 mini", "GPT 5.4 Mini"], "verdict": [{"text": "HN commenters like it as a fast, token-efficient model for implementing coding features", "cites": [{"id": 776437424, "item_id": 48527339, "url": "https://news.ycombinator.com/item?id=48527339", "by": "zhshshshs", "date": "2026-06-14", "stance": "recommend", "quote": "I use 5.4 mini all time for coding. No trouble letting it implement features.", "excerpt": "Counter: I use 5.4 mini all time for coding. No trouble letting it implement features. Entire new screens, APIs and various components. It ain\u2019t the best for sure, but if you have trouble letting it move 20 lines I don\u2019t know what\u2019s the cause but that\u2019s not my experience at all. I do make pretty extensive use of guardrails and proper instructions in my AGENTS.md. I also value super boring code bases with an as much a", "context": "", "use_case": "coding features", "switched_from": "", "story": "GLM 5.2 Is Out"}, {"id": 774466422, "item_id": 48404151, "url": "https://news.ycombinator.com/item?id=48404151", "by": "phsource", "date": "2026-06-04", "stance": "recommend", "quote": "Kimi K2.5, Composer 2.5, GPT-5.4 mini, etc. for faster implementation", "excerpt": "Personally, with our company on Cursor, I can see why model makers are not the best people to go all the way down the stack. Using the right model for the situation will continue to be important, and model makers, by design, do not want to give you the choice to run different models. Right now, we use: - Kimi K2.5 for easy fixes, asking about the code, various agentic commands (e.g., summarizing Loom videos for Slack", "context": "", "use_case": "faster implementation", "switched_from": "", "story": "Show HN: Boxes.dev: ditch localhost; run Claude Code and Codex in the cloud"}, {"id": 762216210, "item_id": 47638513, "url": "https://news.ycombinator.com/item?id=47638513", "by": "syl5x", "date": "2026-04-04", "stance": "recommend", "quote": "I like the ability to start \"tasks\" with faster models like gpt5.4-mini for certain tasks", "excerpt": "I am actively using ohmypi harness which is based on pi-mono which I believe is within OpenClaw, I don't personally use OpenClaw but I suspect that I will be affected. The reason that I use ohmypi is because I can extend it and put guardrails specific for our company and myself (those are different from SKILLs and more sophisticated than the hooks) + I like the ability to start \"tasks\" with faster models like gpt5.4-", "context": "", "use_case": "fast subtasks", "switched_from": "", "story": "Anthropic no longer allowing Claude Code subscriptions to be used with OpenClaw"}, {"id": 793162384, "item_id": 49572649, "url": "https://news.ycombinator.com/item?id=49572649", "by": "shubhamjain", "date": "2026-09-05", "stance": "recommend", "quote": "I switched to GPT-5.4-mini, which uses far fewer tokens, and it\u2019s often just as good as Sonnet.", "excerpt": "> So this is just delegating certain work to dumber models? I certainly wouldn't use Gemini 2.5 Flash (!!?) for code writing as suggested. Why not, though? I started using OpenCode + GitHub Copilot, but I burned through my Claude Sonnet quota in just three days. I switched to GPT-5.4-mini, which uses far fewer tokens, and it\u2019s often just as good as Sonnet. I think optimizing token usage is a good exercise. We often a", "context": "", "use_case": "coding with fewer tokens", "switched_from": "Claude Sonnet", "story": "Portal by Spotify cut my Claude Code token usage by 90%"}]}, {"text": ", and some find it good at code review.", "cites": [{"id": 795271329, "item_id": 49704458, "url": "https://news.ycombinator.com/item?id=49704458", "by": "CharlieDigital", "date": "2026-09-14", "stance": "recommend", "quote": "I found Luna and even 5.4-mini to be quite good at code review", "excerpt": "I found Luna and even 5.4-mini to be quite good at code review provided a few things: 1. Run it in multiple cycles, only on the diff, and only emit a few findings at a time. 2. Give it a memory so each cycle, it knows the previous finding to check if it's been fixed. 3. Give it access to canonical docs that encode your human reviewer heuristics. I exposed these as tool calls so they could be tracked via telemetry. 4.", "context": "", "use_case": "code review", "switched_from": "", "story": "GPT-5.6 Luna vs. GPT-6 Astra: Is a $1.20 Model Good Enough for Code Review?"}]}, {"text": " Those reporting good results rely on guardrails and structured prompts", "cites": [{"id": 776437424, "item_id": 48527339, "url": "https://news.ycombinator.com/item?id=48527339", "by": "zhshshshs", "date": "2026-06-14", "stance": "recommend", "quote": "I use 5.4 mini all time for coding. No trouble letting it implement features.", "excerpt": "Counter: I use 5.4 mini all time for coding. No trouble letting it implement features. Entire new screens, APIs and various components. It ain\u2019t the best for sure, but if you have trouble letting it move 20 lines I don\u2019t know what\u2019s the cause but that\u2019s not my experience at all. I do make pretty extensive use of guardrails and proper instructions in my AGENTS.md. I also value super boring code bases with an as much a", "context": "", "use_case": "coding features", "switched_from": "", "story": "GLM 5.2 Is Out"}, {"id": 795271329, "item_id": 49704458, "url": "https://news.ycombinator.com/item?id=49704458", "by": "CharlieDigital", "date": "2026-09-14", "stance": "recommend", "quote": "I found Luna and even 5.4-mini to be quite good at code review", "excerpt": "I found Luna and even 5.4-mini to be quite good at code review provided a few things: 1. Run it in multiple cycles, only on the diff, and only emit a few findings at a time. 2. Give it a memory so each cycle, it knows the previous finding to check if it's been fixed. 3. Give it access to canonical docs that encode your human reviewer heuristics. I exposed these as tool calls so they could be tracked via telemetry. 4.", "context": "", "use_case": "code review", "switched_from": "", "story": "GPT-5.6 Luna vs. GPT-6 Astra: Is a $1.20 Model Good Enough for Code Review?"}]}, {"text": ", and one commenter groups it with models that produce hallucinated output.", "cites": [{"id": 785947719, "item_id": 49121732, "url": "https://news.ycombinator.com/item?id=49121732", "by": "user43928", "date": "2026-07-31", "stance": "warn", "quote": "\"I love hallucinated garbage\": Sonnet, Qwen 3.6, GPT 5.4 mini, GPT 5.3 Codex", "excerpt": "In my mind there are three tiers: The SOTA: Fable, GPT 5.6 Sol, Opus 5 The \"enterprise admin did not turn on the new models\": Opus 4.8, GPT 5.5 The \"I love hallucinated garbage\": Sonnet, Qwen 3.6, GPT 5.4 mini, GPT 5.3 Codex, etc. Results vary widely", "context": "", "use_case": "", "switched_from": "", "story": "Google fixed more Chrome bugs in June than over the past two years, thanks to AI"}]}], "praise": [{"label": "Coding implementation", "summary": "Used to implement features, screens, APIs and components in coding workflows.", "quotes": [{"id": 776437424, "item_id": 48527339, "url": "https://news.ycombinator.com/item?id=48527339", "by": "zhshshshs", "date": "2026-06-14", "stance": "recommend", "quote": "I use 5.4 mini all time for coding. No trouble letting it implement features.", "excerpt": "Counter: I use 5.4 mini all time for coding. No trouble letting it implement features. Entire new screens, APIs and various components. It ain\u2019t the best for sure, but if you have trouble letting it move 20 lines I don\u2019t know what\u2019s the cause but that\u2019s not my experience at all. I do make pretty extensive use of guardrails and proper instructions in my AGENTS.md. I also value super boring code bases with an as much a", "context": "", "use_case": "coding features", "switched_from": "", "story": "GLM 5.2 Is Out"}, {"id": 774466422, "item_id": 48404151, "url": "https://news.ycombinator.com/item?id=48404151", "by": "phsource", "date": "2026-06-04", "stance": "recommend", "quote": "Kimi K2.5, Composer 2.5, GPT-5.4 mini, etc. for faster implementation", "excerpt": "Personally, with our company on Cursor, I can see why model makers are not the best people to go all the way down the stack. Using the right model for the situation will continue to be important, and model makers, by design, do not want to give you the choice to run different models. Right now, we use: - Kimi K2.5 for easy fixes, asking about the code, various agentic commands (e.g., summarizing Loom videos for Slack", "context": "", "use_case": "faster implementation", "switched_from": "", "story": "Show HN: Boxes.dev: ditch localhost; run Claude Code and Codex in the cloud"}, {"id": 793162384, "item_id": 49572649, "url": "https://news.ycombinator.com/item?id=49572649", "by": "shubhamjain", "date": "2026-09-05", "stance": "recommend", "quote": "I switched to GPT-5.4-mini, which uses far fewer tokens, and it\u2019s often just as good as Sonnet.", "excerpt": "> So this is just delegating certain work to dumber models? I certainly wouldn't use Gemini 2.5 Flash (!!?) for code writing as suggested. Why not, though? I started using OpenCode + GitHub Copilot, but I burned through my Claude Sonnet quota in just three days. I switched to GPT-5.4-mini, which uses far fewer tokens, and it\u2019s often just as good as Sonnet. I think optimizing token usage is a good exercise. We often a", "context": "", "use_case": "coding with fewer tokens", "switched_from": "Claude Sonnet", "story": "Portal by Spotify cut my Claude Code token usage by 90%"}]}, {"label": "Speed", "summary": "Chosen as a faster model for implementation and delegated tasks.", "quotes": [{"id": 762216210, "item_id": 47638513, "url": "https://news.ycombinator.com/item?id=47638513", "by": "syl5x", "date": "2026-04-04", "stance": "recommend", "quote": "I like the ability to start \"tasks\" with faster models like gpt5.4-mini for certain tasks", "excerpt": "I am actively using ohmypi harness which is based on pi-mono which I believe is within OpenClaw, I don't personally use OpenClaw but I suspect that I will be affected. The reason that I use ohmypi is because I can extend it and put guardrails specific for our company and myself (those are different from SKILLs and more sophisticated than the hooks) + I like the ability to start \"tasks\" with faster models like gpt5.4-", "context": "", "use_case": "fast subtasks", "switched_from": "", "story": "Anthropic no longer allowing Claude Code subscriptions to be used with OpenClaw"}, {"id": 774466422, "item_id": 48404151, "url": "https://news.ycombinator.com/item?id=48404151", "by": "phsource", "date": "2026-06-04", "stance": "recommend", "quote": "Kimi K2.5, Composer 2.5, GPT-5.4 mini, etc. for faster implementation", "excerpt": "Personally, with our company on Cursor, I can see why model makers are not the best people to go all the way down the stack. Using the right model for the situation will continue to be important, and model makers, by design, do not want to give you the choice to run different models. Right now, we use: - Kimi K2.5 for easy fixes, asking about the code, various agentic commands (e.g., summarizing Loom videos for Slack", "context": "", "use_case": "faster implementation", "switched_from": "", "story": "Show HN: Boxes.dev: ditch localhost; run Claude Code and Codex in the cloud"}]}, {"label": "Token efficiency", "summary": "Uses far fewer tokens than Sonnet and is often about as good.", "quotes": [{"id": 793162384, "item_id": 49572649, "url": "https://news.ycombinator.com/item?id=49572649", "by": "shubhamjain", "date": "2026-09-05", "stance": "recommend", "quote": "I switched to GPT-5.4-mini, which uses far fewer tokens, and it\u2019s often just as good as Sonnet.", "excerpt": "> So this is just delegating certain work to dumber models? I certainly wouldn't use Gemini 2.5 Flash (!!?) for code writing as suggested. Why not, though? I started using OpenCode + GitHub Copilot, but I burned through my Claude Sonnet quota in just three days. I switched to GPT-5.4-mini, which uses far fewer tokens, and it\u2019s often just as good as Sonnet. I think optimizing token usage is a good exercise. We often a", "context": "", "use_case": "coding with fewer tokens", "switched_from": "Claude Sonnet", "story": "Portal by Spotify cut my Claude Code token usage by 90%"}]}, {"label": "Code review", "summary": "Considered quite good at code review when run in focused, iterative cycles.", "quotes": [{"id": 795271329, "item_id": 49704458, "url": "https://news.ycombinator.com/item?id=49704458", "by": "CharlieDigital", "date": "2026-09-14", "stance": "recommend", "quote": "I found Luna and even 5.4-mini to be quite good at code review", "excerpt": "I found Luna and even 5.4-mini to be quite good at code review provided a few things: 1. Run it in multiple cycles, only on the diff, and only emit a few findings at a time. 2. Give it a memory so each cycle, it knows the previous finding to check if it's been fixed. 3. Give it access to canonical docs that encode your human reviewer heuristics. I exposed these as tool calls so they could be tracked via telemetry. 4.", "context": "", "use_case": "code review", "switched_from": "", "story": "GPT-5.6 Luna vs. GPT-6 Astra: Is a $1.20 Model Good Enough for Code Review?"}]}], "caveats": [{"label": "Needs guardrails", "summary": "Good results depend on guardrails, clear instructions and structured review setups.", "quotes": [{"id": 776437424, "item_id": 48527339, "url": "https://news.ycombinator.com/item?id=48527339", "by": "zhshshshs", "date": "2026-06-14", "stance": "recommend", "quote": "I use 5.4 mini all time for coding. No trouble letting it implement features.", "excerpt": "Counter: I use 5.4 mini all time for coding. No trouble letting it implement features. Entire new screens, APIs and various components. It ain\u2019t the best for sure, but if you have trouble letting it move 20 lines I don\u2019t know what\u2019s the cause but that\u2019s not my experience at all. I do make pretty extensive use of guardrails and proper instructions in my AGENTS.md. I also value super boring code bases with an as much a", "context": "", "use_case": "coding features", "switched_from": "", "story": "GLM 5.2 Is Out"}, {"id": 795271329, "item_id": 49704458, "url": "https://news.ycombinator.com/item?id=49704458", "by": "CharlieDigital", "date": "2026-09-14", "stance": "recommend", "quote": "I found Luna and even 5.4-mini to be quite good at code review", "excerpt": "I found Luna and even 5.4-mini to be quite good at code review provided a few things: 1. Run it in multiple cycles, only on the diff, and only emit a few findings at a time. 2. Give it a memory so each cycle, it knows the previous finding to check if it's been fixed. 3. Give it access to canonical docs that encode your human reviewer heuristics. I exposed these as tool calls so they could be tracked via telemetry. 4.", "context": "", "use_case": "code review", "switched_from": "", "story": "GPT-5.6 Luna vs. GPT-6 Astra: Is a $1.20 Model Good Enough for Code Review?"}]}, {"label": "Hallucinations", "summary": "One commenter places it in a tier of models that produce hallucinated output.", "quotes": [{"id": 785947719, "item_id": 49121732, "url": "https://news.ycombinator.com/item?id=49121732", "by": "user43928", "date": "2026-07-31", "stance": "warn", "quote": "\"I love hallucinated garbage\": Sonnet, Qwen 3.6, GPT 5.4 mini, GPT 5.3 Codex", "excerpt": "In my mind there are three tiers: The SOTA: Fable, GPT 5.6 Sol, Opus 5 The \"enterprise admin did not turn on the new models\": Opus 4.8, GPT 5.5 The \"I love hallucinated garbage\": Sonnet, Qwen 3.6, GPT 5.4 mini, GPT 5.3 Codex, etc. Results vary widely", "context": "", "use_case": "", "switched_from": "", "story": "Google fixed more Chrome bugs in June than over the past two years, thanks to AI"}]}, {"label": "Not top-tier", "summary": "Even supporters say it is not the best model available.", "quotes": [{"id": 776437424, "item_id": 48527339, "url": "https://news.ycombinator.com/item?id=48527339", "by": "zhshshshs", "date": "2026-06-14", "stance": "recommend", "quote": "I use 5.4 mini all time for coding. No trouble letting it implement features.", "excerpt": "Counter: I use 5.4 mini all time for coding. No trouble letting it implement features. Entire new screens, APIs and various components. It ain\u2019t the best for sure, but if you have trouble letting it move 20 lines I don\u2019t know what\u2019s the cause but that\u2019s not my experience at all. I do make pretty extensive use of guardrails and proper instructions in my AGENTS.md. I also value super boring code bases with an as much a", "context": "", "use_case": "coding features", "switched_from": "", "story": "GLM 5.2 Is Out"}, {"id": 785947719, "item_id": 49121732, "url": "https://news.ycombinator.com/item?id=49121732", "by": "user43928", "date": "2026-07-31", "stance": "warn", "quote": "\"I love hallucinated garbage\": Sonnet, Qwen 3.6, GPT 5.4 mini, GPT 5.3 Codex", "excerpt": "In my mind there are three tiers: The SOTA: Fable, GPT 5.6 Sol, Opus 5 The \"enterprise admin did not turn on the new models\": Opus 4.8, GPT 5.5 The \"I love hallucinated garbage\": Sonnet, Qwen 3.6, GPT 5.4 mini, GPT 5.3 Codex, etc. Results vary widely", "context": "", "use_case": "", "switched_from": "", "story": "Google fixed more Chrome bugs in June than over the past two years, thanks to AI"}]}], "who": [], "tags": ["for coding agents", "for code review", "for limited token budgets", "for multi-model setups"], "use_cases": ["code review", "coding with fewer tokens", "coding features", "faster implementation", "fast subtasks"], "compared": [{"slug": "claude-sonnet", "name": "Claude Sonnet", "co": 3, "switched_from": 1, "switched_to": 0}, {"slug": "gpt-5", "name": "GPT-5", "co": 4, "switched_from": 0, "switched_to": 0}, {"slug": "claude-opus", "name": "Claude Opus", "co": 3, "switched_from": 0, "switched_to": 0}, {"slug": "oh-my-pi", "name": "oh-my-pi", "co": 2, "switched_from": 0, "switched_to": 0}, {"slug": "opencode", "name": "OpenCode", "co": 2, "switched_from": 0, "switched_to": 0}], "makers": [], "threads": [{"url": "https://news.ycombinator.com/item?id=49703003", "title": "GPT-5.6 Luna vs. GPT-6 Astra: Is a $1.20 Model Good Enough for Code Review?", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=49571465", "title": "Portal by Spotify cut my Claude Code token usage by 90%", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=49120097", "title": "Google fixed more Chrome bugs in June than over the past two years, thanks to AI", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=48518684", "title": "GLM 5.2 Is Out", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=48399358", "title": "Show HN: Boxes.dev: ditch localhost; run Claude Code and Codex in the cloud", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=47633396", "title": "Anthropic no longer allowing Claude Code subscriptions to be used with OpenClaw", "mentions": 1}], "sentiment_by_month": {"2025-11": {}, "2025-12": {}, "2026-01": {}, "2026-02": {}, "2026-03": {}, "2026-04": {"recommend": 1}, "2026-05": {}, "2026-06": {"recommend": 2}, "2026-07": {"warn": 1}, "2026-08": {}, "2026-09": {"recommend": 2}, "2026-10": {}}, "quotes": [{"id": 795271329, "item_id": 49704458, "url": "https://news.ycombinator.com/item?id=49704458", "by": "CharlieDigital", "date": "2026-09-14", "stance": "recommend", "quote": "I found Luna and even 5.4-mini to be quite good at code review", "excerpt": "I found Luna and even 5.4-mini to be quite good at code review provided a few things: 1. Run it in multiple cycles, only on the diff, and only emit a few findings at a time. 2. Give it a memory so each cycle, it knows the previous finding to check if it's been fixed. 3. Give it access to canonical docs that encode your human reviewer heuristics. I exposed these as tool calls so they could be tracked via telemetry. 4.", "context": "", "use_case": "code review", "switched_from": "", "story": "GPT-5.6 Luna vs. GPT-6 Astra: Is a $1.20 Model Good Enough for Code Review?"}, {"id": 793162384, "item_id": 49572649, "url": "https://news.ycombinator.com/item?id=49572649", "by": "shubhamjain", "date": "2026-09-05", "stance": "recommend", "quote": "I switched to GPT-5.4-mini, which uses far fewer tokens, and it\u2019s often just as good as Sonnet.", "excerpt": "> So this is just delegating certain work to dumber models? I certainly wouldn't use Gemini 2.5 Flash (!!?) for code writing as suggested. Why not, though? I started using OpenCode + GitHub Copilot, but I burned through my Claude Sonnet quota in just three days. I switched to GPT-5.4-mini, which uses far fewer tokens, and it\u2019s often just as good as Sonnet. I think optimizing token usage is a good exercise. We often a", "context": "", "use_case": "coding with fewer tokens", "switched_from": "Claude Sonnet", "story": "Portal by Spotify cut my Claude Code token usage by 90%"}, {"id": 785947719, "item_id": 49121732, "url": "https://news.ycombinator.com/item?id=49121732", "by": "user43928", "date": "2026-07-31", "stance": "warn", "quote": "\"I love hallucinated garbage\": Sonnet, Qwen 3.6, GPT 5.4 mini, GPT 5.3 Codex", "excerpt": "In my mind there are three tiers: The SOTA: Fable, GPT 5.6 Sol, Opus 5 The \"enterprise admin did not turn on the new models\": Opus 4.8, GPT 5.5 The \"I love hallucinated garbage\": Sonnet, Qwen 3.6, GPT 5.4 mini, GPT 5.3 Codex, etc. Results vary widely", "context": "", "use_case": "", "switched_from": "", "story": "Google fixed more Chrome bugs in June than over the past two years, thanks to AI"}, {"id": 776437424, "item_id": 48527339, "url": "https://news.ycombinator.com/item?id=48527339", "by": "zhshshshs", "date": "2026-06-14", "stance": "recommend", "quote": "I use 5.4 mini all time for coding. No trouble letting it implement features.", "excerpt": "Counter: I use 5.4 mini all time for coding. No trouble letting it implement features. Entire new screens, APIs and various components. It ain\u2019t the best for sure, but if you have trouble letting it move 20 lines I don\u2019t know what\u2019s the cause but that\u2019s not my experience at all. I do make pretty extensive use of guardrails and proper instructions in my AGENTS.md. I also value super boring code bases with an as much a", "context": "", "use_case": "coding features", "switched_from": "", "story": "GLM 5.2 Is Out"}, {"id": 774466422, "item_id": 48404151, "url": "https://news.ycombinator.com/item?id=48404151", "by": "phsource", "date": "2026-06-04", "stance": "recommend", "quote": "Kimi K2.5, Composer 2.5, GPT-5.4 mini, etc. for faster implementation", "excerpt": "Personally, with our company on Cursor, I can see why model makers are not the best people to go all the way down the stack. Using the right model for the situation will continue to be important, and model makers, by design, do not want to give you the choice to run different models. Right now, we use: - Kimi K2.5 for easy fixes, asking about the code, various agentic commands (e.g., summarizing Loom videos for Slack", "context": "", "use_case": "faster implementation", "switched_from": "", "story": "Show HN: Boxes.dev: ditch localhost; run Claude Code and Codex in the cloud"}, {"id": 762216210, "item_id": 47638513, "url": "https://news.ycombinator.com/item?id=47638513", "by": "syl5x", "date": "2026-04-04", "stance": "recommend", "quote": "I like the ability to start \"tasks\" with faster models like gpt5.4-mini for certain tasks", "excerpt": "I am actively using ohmypi harness which is based on pi-mono which I believe is within OpenClaw, I don't personally use OpenClaw but I suspect that I will be affected. The reason that I use ohmypi is because I can extend it and put guardrails specific for our company and myself (those are different from SKILLs and more sophisticated than the hooks) + I like the ability to start \"tasks\" with faster models like gpt5.4-", "context": "", "use_case": "fast subtasks", "switched_from": "", "story": "Anthropic no longer allowing Claude Code subscriptions to be used with OpenClaw"}]}