OpenAI compatible API · Attested · Public status

Compare AI models

Compare context, provider routes, pricing, privacy posture, and measured performance across the TrustedRouter model catalog.

Verify gateway

Onebase URL to migrate

100sof models and routes

0prompt or output logs. Always.

Measured speed

Find the fastest routes.

Compare provider TTFT, throughput, and uptime from routed probes.

Open leaderboard → Open weights

Compare open models.

Browse DeepSeek, GLM, Gemma, Kimi, MiniMax, Qwen, and other open routes.

Browse models → Privacy

Require ZDR or confidential routes.

Separate gateway attestation from each downstream provider's privacy posture.

Review privacy →

Current comparison, preserved evidence.

Comparison pages use the current catalog and live route measurements. Monthly reports preserve the historical evidence after prices and provider performance change.

Monthly reports Run your own eval

Model comparisons

Pick two models. Compare the routes that actually serve them.

Every comparison uses the current TrustedRouter catalog, prices, provider posture, and measured route data.

Comparison	First model	Routes	Second model	Routes
OpenAI: gpt-oss-20b vs Qwen3 235B A22B Instruct 2507	`openai/gpt-oss-20b`	18	`qwen/qwen3-235b-a22b-instruct-2507`	9
OpenAI: gpt-oss-20b vs Xiaomi: MiMo-V2.5	`openai/gpt-oss-20b`	18	`xiaomi/mimo-v2.5`	8
DeepSeek: DeepSeek V4 Flash Vision Exp vs OpenAI: gpt-oss-20b	`deepseek/deepseek-v4-flash-vision-exp`	10	`openai/gpt-oss-20b`	18
Google: Gemini 2.5 Flash vs OpenAI: gpt-oss-20b	`google/gemini-2.5-flash`	8	`openai/gpt-oss-20b`	18
NVIDIA: Nemotron 3 Ultra vs OpenAI: gpt-oss-20b	`nvidia/nemotron-3-ultra-550b-a55b`	9	`openai/gpt-oss-20b`	18
OpenAI: gpt-oss-20b vs Qwen: Qwen3.5-9B	`openai/gpt-oss-20b`	18	`qwen/qwen3.5-9b`	10
OpenAI: gpt-oss-20b vs Z.ai: GLM 4.6	`openai/gpt-oss-20b`	18	`z-ai/glm-4.6`	8
DeepSeek V3.1 vs OpenAI: gpt-oss-20b	`deepseek/deepseek-v3.1`	7	`openai/gpt-oss-20b`	18
OpenAI: gpt-oss-20b vs Xiaomi: MiMo-V2.5-Pro	`openai/gpt-oss-20b`	18	`xiaomi/mimo-v2.5-pro`	8
DeepSeek V4 Pro 0423 vs OpenAI: gpt-oss-20b	`deepseek/deepseek-v4-pro-0423`	4	`openai/gpt-oss-20b`	18
Google: Gemini 2.5 Pro vs OpenAI: gpt-oss-20b	`google/gemini-2.5-pro`	6	`openai/gpt-oss-20b`	18
Google: Gemini 3 Flash Preview vs OpenAI: gpt-oss-20b	`google/gemini-3-flash-preview`	6	`openai/gpt-oss-20b`	18
Google: Gemini 3.1 Flash Lite vs OpenAI: gpt-oss-20b	`google/gemini-3.1-flash-lite`	6	`openai/gpt-oss-20b`	18
Google: Gemini 3.1 Pro Preview vs OpenAI: gpt-oss-20b	`google/gemini-3.1-pro-preview`	7	`openai/gpt-oss-20b`	18
Google: Gemini 3.5 Flash vs OpenAI: gpt-oss-20b	`google/gemini-3.5-flash`	7	`openai/gpt-oss-20b`	18
OpenAI: gpt-oss-20b vs Qwen3.7 Max	`openai/gpt-oss-20b`	18	`qwen/qwen3.7-max`	7
OpenAI: gpt-oss-20b vs Qwen3.8 Max	`openai/gpt-oss-20b`	18	`qwen/qwen3.8-max`	8
OpenAI: gpt-oss-20b vs Thinking Machines: Inkling	`openai/gpt-oss-20b`	18	`thinkingmachines/inkling`	7
OpenAI: gpt-oss-20b vs Thinking Machines: Inkling Small	`openai/gpt-oss-20b`	18	`thinkingmachines/inkling-small`	7
NVIDIA: Nemotron 3 Super vs OpenAI: gpt-oss-20b	`nvidia/nemotron-3-super-120b-a12b`	7	`openai/gpt-oss-20b`	18
OpenAI: gpt-oss-20b vs Qwen: Qwen3 Coder Next	`openai/gpt-oss-20b`	18	`qwen/qwen3-coder-next`	7
OpenAI: gpt-oss-20b vs Qwen: Qwen3 VL 235B A22B Instruct	`openai/gpt-oss-20b`	18	`qwen/qwen3-vl-235b-a22b-instruct`	7
OpenAI: gpt-oss-20b vs Qwen: Qwen3.5-27B	`openai/gpt-oss-20b`	18	`qwen/qwen3.5-27b`	6
OpenAI: gpt-oss-20b vs Qwen: Qwen3.8 2.4T A95B	`openai/gpt-oss-20b`	18	`qwen/qwen3.8-2.4t-a95b`	8
OpenAI: gpt-oss-20b vs Tencent: Hy3	`openai/gpt-oss-20b`	18	`tencent/hy3`	8
Google: Gemma 3 27B vs OpenAI: gpt-oss-20b	`google/gemma-3-27b-it`	8	`openai/gpt-oss-20b`	18
Meta: Llama 3.1 8B Instruct vs OpenAI: gpt-oss-20b	`meta-llama/llama-3.1-8b-instruct`	6	`openai/gpt-oss-20b`	18
OpenAI: gpt-oss-20b vs Qwen: Qwen3 235B A22B Thinking 2507	`openai/gpt-oss-20b`	18	`qwen/qwen3-235b-a22b-thinking-2507`	7
OpenAI: GPT-5.6 Terra vs OpenAI: gpt-oss-20b	`openai/gpt-5.6-terra`	6	`openai/gpt-oss-20b`	18
OpenAI: GPT-6 Astra vs OpenAI: gpt-oss-20b	`openai/gpt-6-astra`	6	`openai/gpt-oss-20b`	18
Llama 4 Maverick Instruct vs OpenAI: gpt-oss-20b	`meta-llama/llama-4-maverick-17b-128e-instruct-fp8`	5	`openai/gpt-oss-20b`	18
NVIDIA: Nemotron 3 Nano 30B A3B vs OpenAI: gpt-oss-20b	`nvidia/nemotron-3-nano-30b-a3b`	6	`openai/gpt-oss-20b`	18
nvidia/Nemotron-3.5-Lightning-30B-A3B vs OpenAI: gpt-oss-20b	`nvidia/nemotron-3.5-lightning-30b-a3b`	4	`openai/gpt-oss-20b`	18
OpenAI: gpt-oss-20b vs Qwen3 Max	`openai/gpt-oss-20b`	18	`qwen/qwen3-max`	5
OpenAI: gpt-oss-20b vs Qwen: Qwen3 VL 30B A3B Instruct	`openai/gpt-oss-20b`	18	`qwen/qwen3-vl-30b-a3b-instruct`	5
OpenAI: gpt-oss-20b vs Qwen: Qwen3 VL 8B Instruct	`openai/gpt-oss-20b`	18	`qwen/qwen3-vl-8b-instruct`	4
OpenAI: gpt-oss-20b vs Z.ai: GLM 5V Turbo	`openai/gpt-oss-20b`	18	`z-ai/glm-5v-turbo`	5
DeepSeek: R1 0528 vs OpenAI: gpt-oss-20b	`deepseek/deepseek-r1-0528`	6	`openai/gpt-oss-20b`	18
OpenAI: gpt-oss-20b vs Qwen: Qwen3 Coder 30B A3B Instruct	`openai/gpt-oss-20b`	18	`qwen/qwen3-coder-30b-a3b-instruct`	4
Google: Gemma 3 12B vs OpenAI: gpt-oss-20b	`google/gemma-3-12b-it`	5	`openai/gpt-oss-20b`	18
Meta: Llama 3.1 70B Instruct vs OpenAI: gpt-oss-20b	`meta-llama/llama-3.1-70b-instruct`	5	`openai/gpt-oss-20b`	18
Llama 4 Scout Instruct vs OpenAI: gpt-oss-20b	`meta-llama/llama-4-scout-17b-16e-instruct`	5	`openai/gpt-oss-20b`	18
Muse Glimmer 30B on Fireworks vs OpenAI: gpt-oss-20b	`meta-models/muse-glimmer-30b`	6	`openai/gpt-oss-20b`	18
Google: Gemma 4 26B A4B vs Qwen: Qwen3.5 397B A17B	`google/gemma-4-26b-a4b-it`	15	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3.5 397B A17B vs Qwen: Qwen3.6 27B	`qwen/qwen3.5-397b-a17b`	17	`qwen/qwen3.6-27b`	14
DeepSeek: DeepSeek V3.2 vs Qwen: Qwen3.5 397B A17B	`deepseek/deepseek-v3.2`	16	`qwen/qwen3.5-397b-a17b`	17
MoonshotAI: Kimi K2.5 vs Qwen: Qwen3.5 397B A17B	`moonshotai/kimi-k2.5`	12	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3.5-122B-A10B vs Qwen: Qwen3.5 397B A17B	`qwen/qwen3.5-122b-a10b`	12	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3.5 397B A17B vs Z.ai: GLM 4.7	`qwen/qwen3.5-397b-a17b`	17	`z-ai/glm-4.7`	11
MiniMax: MiniMax M2.5 vs Qwen: Qwen3.5 397B A17B	`minimax/minimax-m2.5`	13	`qwen/qwen3.5-397b-a17b`	17
MiniMax: MiniMax M2.7 vs Qwen: Qwen3.5 397B A17B	`minimax/minimax-m2.7`	11	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3 Next 80B A3B Instruct vs Qwen: Qwen3.5 397B A17B	`qwen/qwen3-next-80b-a3b-instruct`	10	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3.5-35B-A3B vs Qwen: Qwen3.5 397B A17B	`qwen/qwen3.5-35b-a3b`	10	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3.5 397B A17B vs Z.ai: GLM 5	`qwen/qwen3.5-397b-a17b`	17	`z-ai/glm-5`	10
Qwen3 235B A22B Instruct 2507 vs Qwen: Qwen3.5 397B A17B	`qwen/qwen3-235b-a22b-instruct-2507`	9	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3.5 397B A17B vs Xiaomi: MiMo-V2.5	`qwen/qwen3.5-397b-a17b`	17	`xiaomi/mimo-v2.5`	8
DeepSeek: DeepSeek V4 Flash Vision Exp vs Qwen: Qwen3.5 397B A17B	`deepseek/deepseek-v4-flash-vision-exp`	10	`qwen/qwen3.5-397b-a17b`	17
Google: Gemini 2.5 Flash vs Qwen: Qwen3.5 397B A17B	`google/gemini-2.5-flash`	8	`qwen/qwen3.5-397b-a17b`	17
NVIDIA: Nemotron 3 Ultra vs Qwen: Qwen3.5 397B A17B	`nvidia/nemotron-3-ultra-550b-a55b`	9	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3.5 397B A17B vs Qwen: Qwen3.5-9B	`qwen/qwen3.5-397b-a17b`	17	`qwen/qwen3.5-9b`	10
Qwen: Qwen3.5 397B A17B vs Z.ai: GLM 4.6	`qwen/qwen3.5-397b-a17b`	17	`z-ai/glm-4.6`	8
DeepSeek V3.1 vs Qwen: Qwen3.5 397B A17B	`deepseek/deepseek-v3.1`	7	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3.5 397B A17B vs Xiaomi: MiMo-V2.5-Pro	`qwen/qwen3.5-397b-a17b`	17	`xiaomi/mimo-v2.5-pro`	8
DeepSeek V4 Pro 0423 vs Qwen: Qwen3.5 397B A17B	`deepseek/deepseek-v4-pro-0423`	4	`qwen/qwen3.5-397b-a17b`	17
Google: Gemini 2.5 Pro vs Qwen: Qwen3.5 397B A17B	`google/gemini-2.5-pro`	6	`qwen/qwen3.5-397b-a17b`	17
Google: Gemini 3 Flash Preview vs Qwen: Qwen3.5 397B A17B	`google/gemini-3-flash-preview`	6	`qwen/qwen3.5-397b-a17b`	17
Google: Gemini 3.1 Flash Lite vs Qwen: Qwen3.5 397B A17B	`google/gemini-3.1-flash-lite`	6	`qwen/qwen3.5-397b-a17b`	17
Google: Gemini 3.1 Pro Preview vs Qwen: Qwen3.5 397B A17B	`google/gemini-3.1-pro-preview`	7	`qwen/qwen3.5-397b-a17b`	17
Google: Gemini 3.5 Flash vs Qwen: Qwen3.5 397B A17B	`google/gemini-3.5-flash`	7	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3.5 397B A17B vs Qwen3.7 Max	`qwen/qwen3.5-397b-a17b`	17	`qwen/qwen3.7-max`	7
Qwen: Qwen3.5 397B A17B vs Qwen3.8 Max	`qwen/qwen3.5-397b-a17b`	17	`qwen/qwen3.8-max`	8
Qwen: Qwen3.5 397B A17B vs Thinking Machines: Inkling	`qwen/qwen3.5-397b-a17b`	17	`thinkingmachines/inkling`	7
Qwen: Qwen3.5 397B A17B vs Thinking Machines: Inkling Small	`qwen/qwen3.5-397b-a17b`	17	`thinkingmachines/inkling-small`	7
NVIDIA: Nemotron 3 Super vs Qwen: Qwen3.5 397B A17B	`nvidia/nemotron-3-super-120b-a12b`	7	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3 Coder Next vs Qwen: Qwen3.5 397B A17B	`qwen/qwen3-coder-next`	7	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3 VL 235B A22B Instruct vs Qwen: Qwen3.5 397B A17B	`qwen/qwen3-vl-235b-a22b-instruct`	7	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3.5-27B vs Qwen: Qwen3.5 397B A17B	`qwen/qwen3.5-27b`	6	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3.5 397B A17B vs Qwen: Qwen3.8 2.4T A95B	`qwen/qwen3.5-397b-a17b`	17	`qwen/qwen3.8-2.4t-a95b`	8
Qwen: Qwen3.5 397B A17B vs Tencent: Hy3	`qwen/qwen3.5-397b-a17b`	17	`tencent/hy3`	8
Google: Gemma 3 27B vs Qwen: Qwen3.5 397B A17B	`google/gemma-3-27b-it`	8	`qwen/qwen3.5-397b-a17b`	17
Meta: Llama 3.1 8B Instruct vs Qwen: Qwen3.5 397B A17B	`meta-llama/llama-3.1-8b-instruct`	6	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3 235B A22B Thinking 2507 vs Qwen: Qwen3.5 397B A17B	`qwen/qwen3-235b-a22b-thinking-2507`	7	`qwen/qwen3.5-397b-a17b`	17
OpenAI: GPT-5.6 Terra vs Qwen: Qwen3.5 397B A17B	`openai/gpt-5.6-terra`	6	`qwen/qwen3.5-397b-a17b`	17
OpenAI: GPT-6 Astra vs Qwen: Qwen3.5 397B A17B	`openai/gpt-6-astra`	6	`qwen/qwen3.5-397b-a17b`	17
Llama 4 Maverick Instruct vs Qwen: Qwen3.5 397B A17B	`meta-llama/llama-4-maverick-17b-128e-instruct-fp8`	5	`qwen/qwen3.5-397b-a17b`	17
NVIDIA: Nemotron 3 Nano 30B A3B vs Qwen: Qwen3.5 397B A17B	`nvidia/nemotron-3-nano-30b-a3b`	6	`qwen/qwen3.5-397b-a17b`	17
nvidia/Nemotron-3.5-Lightning-30B-A3B vs Qwen: Qwen3.5 397B A17B	`nvidia/nemotron-3.5-lightning-30b-a3b`	4	`qwen/qwen3.5-397b-a17b`	17
Qwen3 Max vs Qwen: Qwen3.5 397B A17B	`qwen/qwen3-max`	5	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3 VL 30B A3B Instruct vs Qwen: Qwen3.5 397B A17B	`qwen/qwen3-vl-30b-a3b-instruct`	5	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3 VL 8B Instruct vs Qwen: Qwen3.5 397B A17B	`qwen/qwen3-vl-8b-instruct`	4	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3.5 397B A17B vs Z.ai: GLM 5V Turbo	`qwen/qwen3.5-397b-a17b`	17	`z-ai/glm-5v-turbo`	5
DeepSeek: R1 0528 vs Qwen: Qwen3.5 397B A17B	`deepseek/deepseek-r1-0528`	6	`qwen/qwen3.5-397b-a17b`	17
Qwen: Qwen3 Coder 30B A3B Instruct vs Qwen: Qwen3.5 397B A17B	`qwen/qwen3-coder-30b-a3b-instruct`	4	`qwen/qwen3.5-397b-a17b`	17
Google: Gemma 3 12B vs Qwen: Qwen3.5 397B A17B	`google/gemma-3-12b-it`	5	`qwen/qwen3.5-397b-a17b`	17
Meta: Llama 3.1 70B Instruct vs Qwen: Qwen3.5 397B A17B	`meta-llama/llama-3.1-70b-instruct`	5	`qwen/qwen3.5-397b-a17b`	17
Llama 4 Scout Instruct vs Qwen: Qwen3.5 397B A17B	`meta-llama/llama-4-scout-17b-16e-instruct`	5	`qwen/qwen3.5-397b-a17b`	17
Muse Glimmer 30B on Fireworks vs Qwen: Qwen3.5 397B A17B	`meta-models/muse-glimmer-30b`	6	`qwen/qwen3.5-397b-a17b`	17
Google: Gemma 4 26B A4B vs Qwen: Qwen3.6 27B	`google/gemma-4-26b-a4b-it`	15	`qwen/qwen3.6-27b`	14
DeepSeek: DeepSeek V3.2 vs Google: Gemma 4 26B A4B	`deepseek/deepseek-v3.2`	16	`google/gemma-4-26b-a4b-it`	15
Google: Gemma 4 26B A4B vs MoonshotAI: Kimi K2.5	`google/gemma-4-26b-a4b-it`	15	`moonshotai/kimi-k2.5`	12