Structured-output index
The same model breaks your pipeline's structured outputs 15× more often on some inference providers than on others. Find the best one for your pipelines.
Give this to your agent ↓
← All models
DeepSeek V4.1 Flash 1017 DeepSeek V4.1 Flash DeepSeek V4 Pro 0813 DeepSeek V4 Flash 0731 DeepSeek V4 Pro 0423 DeepSeek V4 Flash 0423 DeepSeek V3.2 DeepSeek V3.1 Gemma 4 26B A4B Gemma 4 31B Gemma 3 27B Muse Spark Open Llama 3.3 70B Instruct MiniMax M3.1 Flash MiniMax M3 Mistral Large 4 Mistral Nemo Kimi K3 Kimi K2.6 Kimi K2.5 Nemotron 3.5 Super VL Nemotron 3.5 Lightning gpt-oss-20b gpt-oss-120b Qwen3.8 27B Qwen3.8 2.4T A95B Qwen3.6 35B A3B Qwen3.5-9B Qwen3.5 397B A17B Qwen3 30B A3B Instruct 2507 Qwen3 235B A22B Instruct 2507 Beam 1.0 501B A23B Hy3 MiMo-V2.6-Pro MiMo-V2.6-Flash GLM 5.3 Flash GLM 5.3 GLM 5.2 GLM 5.1 GLM 5 Mistral · 1 provider · 18k structured outputs a day
# Provider Broken Structured outputs $ in / out Tokens/s Uptime 1 Mistral zdr # 10k $0.70 / $2.10 87 99.9% 2 Mistral # 7.6k $0.70 / $2.10 86 99.9%
No recommended settings for Mistral Large 4 Every provider breaks 3% of structured outputs or more.
Give llms.txt to your agent
Using Mistral Large 4? Get updates when its successor drops And which provider to use.
For model makers Broken structured outputs carry your model's name Your users don't see the provider. They see your model failing.
Worst provider 8.07% of Mistral Large 4's structured outputs fail on Mistral.
Best to worst 1× as often on Mistral as on Mistral (zdr), its best provider.
Every day 1k broken structured outputs a day across its 1 provider.
