Skip to content
RankShark

gpt-oss-20b

gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license.

ModelOpen sourceby OpenAIsince 2025
41.3

SharkScore

#115 in Models

Why it scores 41.3

gpt-oss-20b has a SharkScore of 41.3/100, ranked #115 of 608 in Models (updated 2026-11-19).

Each part compares gpt-oss-20b with other listings in Models. Evidence covers 87% of the score's weight.

Adoption
46 · 30%
User reviews
no data · 25%
Maintenance
1 · 20%
Trust & transparency
80 · 15%
Momentum
no data · 10%

Paying never changes this score. How ranking works · method v2026-11-18.2

RankShark review of gpt-oss-20b

Automated evaluation by RankShark · 18 Nov 2026 · method v1.0 · How we review

We were not paid for this review.

As of 2026-11-18, gpt-oss-20b is an open-weight, text-only 21B model; OpenRouter lists it at $0.018/M input and $0.09/M output tokens. [E1, E2]

OpenAI’s open-weight text-generation model, with 21B parameters and 3.6B active parameters. The model card describes it for lower-latency, local, or specialized use. [E1]

Best for

  • Developers seeking a model they can run locally or for specialized use cases. · model card
  • Developers who want to adjust reasoning effort or fine-tune the model. · model card
  • Applications using text input and output. · OpenRouter

Not for

  • Applications requiring image, audio, or other non-text model inputs or outputs; the listing specifies text input and output. · OpenRouter

Key features

  • Open-weight model. The vendor says gpt-oss-20b is released under the Apache 2.0 license. · model card
  • Reasoning effort settings. The vendor says reasoning effort can be set to low, medium, or high. · model card
  • Fine-tuning. The vendor says the model can be fine-tuned for specialized use cases, including on consumer hardware. · model card
  • Tool use. The vendor says the model supports function calling, web browsing, Python code execution, and Structured Outputs. · model card
  • Local memory requirement. The vendor says gpt-oss-20b runs within 16GB of memory with MXFP4 quantization. · model card
  • Context window. OpenRouter lists a 131,072-token context window. · OpenRouter

Pros

  • Apache 2.0 licensing is stated on the model card. · model card
  • The model card lists multiple ways to run it, including Transformers, vLLM, Ollama, and LM Studio. · model card
  • The vendor says reasoning effort is configurable and the model can be fine-tuned. · model card

Cons

  • The vendor says the model should only be used with the harmony format; it will not work correctly otherwise. · model card
  • The listing specifies text input and output, not multimodal input or output. · OpenRouter
  • No free plan or free trial stated on the pages we checked.
  • Whether customer data is used for model training is not stated on the pages we checked.

Pricing

PlanPriceLimits
OpenRouter inference$0.018 per 1M input tokens; $0.09 per 1M output tokensToken-based rates; monthly plan price or minimum not stated.

Checked 18 Nov 2026

Free plan
Not stated (checked 18 Nov 2026)
Free trial
Not stated (checked 18 Nov 2026)
Trains on customer data
Not stated (checked 18 Nov 2026)
Data retention
Not stated (checked 18 Nov 2026)
Data regions
Not stated (checked 18 Nov 2026)
Data processing agreement
Not stated (checked 18 Nov 2026)
AI models used
gpt-oss-20b; 21B parameters and 3.6B active parameters. · model card
API
OpenRouter lists /api/v1/chat/completions, /api/v1/completions, /api/v1/messages, and /api/v1/responses. · OpenRouter
SDKs
Not stated (checked 18 Nov 2026)
Webhooks
Not stated (checked 18 Nov 2026)
Platforms
The model card provides instructions for Transformers, vLLM, PyTorch/Triton, Ollama, and LM Studio. · model card
Security claims
No security certification stated (checked 18 Nov 2026)

What we measured

Context window (OpenRouter listing) · now
131,072 tokens
Price per million tokens (OpenRouter listing) · now
$0.018 in, $0.09 out
Package downloads · 30 days
5,596,946
How we tested · sources
  • Hourly uptime and response-time checks of the product's URL.
  • Repository, release and download data from GitHub, npm, PyPI and Hugging Face where they exist.
  • We did not sign up, pay, or use the product as a customer.
  • 18 Nov 2026: Plans or prices changed on the pricing page.
  • 18 Nov 2026: Plans or prices changed on the pricing page.

Something wrong? Report an error with a public link that shows it; corrections re-run the review.

Top alternatives to gpt-oss-20b

See all
  • 67.1

    Qwen3.8 27B is an open-weight dense vision-language model from Qwen.

    ModelOpen source
  • 65.5

    Qwen3.6-35B-A3B is an open-weight multimodal model from Alibaba Cloud with 35 billion total parameters and 3 billion active parameters per t

    ModelOpen source
  • 64.7

    Qwen3.8 Flash is a multimodal reasoning model from Alibaba.

    ModelOpen source
  • 64.3

    Qwen3.6 27B is a dense 27-billion-parameter language model from the Qwen Team at Alibaba, released in April 2026.

    ModelOpen source