# Open-weight models across inference providers

24 of the 82 provider endpoints serving 6 open-weight models declare at least one material setting that differs from the official reference (29.3%); 26 of the 71 whose quantization could be compared declare none, so it cannot be checked.

- Endpoints with any declared difference: 32; material: 24
- By kind: 17 context, 5 quantization, 16 tools, 0 reasoning, 5 output-cap
- First read 2026-10-19, last read 2026-10-19
- Every row as JSON: https://www.neurarch.com/providers/index.json

## Models

- openai/gpt-oss-120b: 6 of 24 endpoints with a material difference. https://www.neurarch.com/providers/openai/gpt-oss-120b.html
- openai/gpt-oss-20b: 2 of 11 endpoints with a material difference. https://www.neurarch.com/providers/openai/gpt-oss-20b.html
- moonshotai/kimi-k2.6: 3 of 17 endpoints with a material difference. https://www.neurarch.com/providers/moonshotai/kimi-k2.6.html
- qwen/qwen3-235b-a22b-2507: 3 of 7 endpoints with a material difference. https://www.neurarch.com/providers/qwen/qwen3-235b-a22b-2507.html
- deepseek/deepseek-v3.2: 5 of 12 endpoints with a material difference. https://www.neurarch.com/providers/deepseek/deepseek-v3.2.html
- meta-llama/llama-3.3-70b-instruct: 5 of 11 endpoints with a material difference. https://www.neurarch.com/providers/meta-llama/llama-3.3-70b-instruct.html

## Recent changes

None yet.

## How this is read

Once a day the tracker reads OpenRouter's public endpoint list for each model (one row per provider deployment: declared context length, declared maximum output, declared quantization, and the request parameters it accepts) and the official repository's config.json at its head commit. Both reads are anonymous. A new row is recorded only when either moved, so the history below is a list of real changes.

Every value in the tables is what a provider declares to OpenRouter. No request is sent to any model, so nothing here says how an endpoint behaves: one that declares a 131,072-token context may still cut off earlier, and one that declares the official quantization may still serve something else.

A quantization the provider does not declare cannot be compared, and is counted as undisclosed rather than as matching. A quantization of higher precision than the official weights is shown and not counted as a difference.

Material means one of: a declared context under 95% of the reference config, a declared quantization other than the official release, no tools parameter on a model that supports tool calling, no reasoning control on a model that exposes one, or a reasoning model capped under 8,192 output tokens. A context between 95% and 100% of the reference is listed and not counted as material.

Sampling defaults, chat templates and stop tokens are not in the endpoint metadata and are not compared here. Moonshot's K2 Vendor Verifier (github.com/MoonshotAI/K2-Vendor-Verifier) measures tool-call behaviour across providers for Kimi K2 by sending requests; this page does not.

OpenRouter lists only the providers that serve through it. Nothing here ranks providers, and a provider with no listed difference has not been shown to match the reference, only to declare nothing that differs.