Models with structured outputs / JSON mode
If your backend expects JSON, the model cannot “improvise prose”. This collection prioritizes models with solid structured outputs / JSON mode for APIs, extractors, and tool loops.
Models in this collection
| Model | Type | Context | Price |
|---|---|---|---|
 GPT-5.4 miniOpenAI | Chat | 272k tokens | $0.8175/M in · $4.9053/M out |
| GPT-4.1 mini (legacy)OpenAI | Chat | 1M tokens | $0.436/M in · $1.7441/M out |
 Claude Sonnet 5Anthropic | Chat | 1M tokens | $3.2702/M in · $16.3509/M out |
 Gemini 2.5 FlashGoogle | Chat | 1M tokens | $0.327/M in · $2.7251/M out |
| GPT-5.5OpenAI | Chat | 272k tokens | $5.4503/M in · $32.7018/M out |
| Claude Haiku 4.5Anthropic | Chat | 200k tokens | $1.0901/M in · $5.4503/M out |
 Mistral LargeMistral | Chat | 128k tokens | $2.1801/M in · $6.5404/M out |
| Gemini 2.5 ProGoogle | Chat | 2M tokens | $1.3626/M in · $10.9006/M out |
Why these models
GPT mini/4.1-mini are the JSON-mode standard; Claude and Gemini for schemas + tools; Mistral Large as an alternative; Haiku/Flash for volume.
Use them with Geek Hub
One OpenAI-compatible base URL and API key. Swap any model id from this list without rewriting your client.
Get an API keyFAQ
- JSON mode = tools?
- No. JSON mode fixes response shape; tools call functions. Many models here do both.
- What if JSON fails?
- Retry with the same schema, lower temperature, or validate and re-prompt. Use CI evals (Testing collection).
- OpenAI response_format?
- If your client uses the OpenAI SDK, point base URL to Geek Hub and use a model id from this list.