Deprecated on Sep 15, 2026. 15 providers still list it at the prices below.
Context
131K
Max output
131K
Serving providers
15
Cheapest input
$0.0145 /1M
On-demand listed rates, OpenRouter routed quotes and Vercel reseller quotes. Enterprise commitments (committed-use discounts, provisioned throughput, negotiated rates) can be materially lower.
Benchmarks
independent evaluations · source: Artificial Analysis and Hugging Face leaderboardsSpeed & latency
Output speed
219 t/s
tokens / second
Time to first token
0.77s
latency
Headline indices
Individual benchmarks · click a row for effort variants
| Benchmark | Best score | Effort | Cost / task | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AA-BriefcaseUndated | 0% | default | — | |||||||||||||||||
| APEX Agents | 1% | default | — | |||||||||||||||||
| Automation Bench | 0.0 | default | — | |||||||||||||||||
| CritPt | 1% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| EvalComputeProxyFrozen Aug 18, 2026 | 200.1 | default | — | |||||||||||||||||
| ||||||||||||||||||||
| GDPval v2Frozen Aug 18, 2026 | 566.6 | default | — | |||||||||||||||||
| ||||||||||||||||||||
| GPQA Diamond | 69% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| Humanity's Last Exam | 11% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| IFBench | 65% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| LiveCodeBench | 78% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| Long-Context Reasoning | 35% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| Mlcr OverallFrozen Sep 9, 2026 | 0.0 | default | — | |||||||||||||||||
| MMLU-ProFrozen Aug 18, 2026 | 75% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| Omniscience | -58.5 | low | — | |||||||||||||||||
| ||||||||||||||||||||
| Omniscience Accuracy | 0.2 | default | — | |||||||||||||||||
| ||||||||||||||||||||
| Omniscience Non Hallucination | 0.1 | low | — | |||||||||||||||||
| ||||||||||||||||||||
| SciCode | 39% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| SWE-bench Verified | 61% | default | — | |||||||||||||||||
| Terminal Bench40 | 0 | default | — | |||||||||||||||||
| ||||||||||||||||||||
| TerminalBench Hard | 11% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| TerminalBench v2.1 | 14% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| τ-bench Banking | 7% | default | — | |||||||||||||||||
| τ²-bench | 60% | default | — | |||||||||||||||||
| ||||||||||||||||||||
Artificial Analysis last published scores for this benchmark on Sep 9, 2026. Scores here are frozen at that date, and models released since are not on it.
We cannot tell whether this benchmark is still being published. Treat these scores as undated.
Artificial Analysis last published scores for this benchmark on Aug 18, 2026. Scores here are frozen at that date, and models released since are not on it.
Pricing by serving provider
standard tier · per 1M tokens · click a provider to expandInput pricing runs from $0.0145 to $0.20 per 1M tokens across 15 providers, so the dearest route costs 1279% more than the cheapest for the same model.
| Serving provider | Input /1M | Output /1M | Endpoints | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DarkbloomCheapest | $0.0145 | $0.07 | 1 | |||||||||||||
Standarddarkbloom/gpt-oss-20b
| ||||||||||||||||
| OpenRouter | $0.018 | $0.09 | 1 | |||||||||||||
Standardopenai/gpt-oss-20b
| ||||||||||||||||
| DeepInfra | $0.03 | $0.14 | 1 | |||||||||||||
Standarddeepinfra/openai/gpt-oss-20b
| ||||||||||||||||
| VercelVercel reseller quote | $0.03 | $0.14 | 1 | |||||||||||||
Standardopenai/gpt-oss-20b
| ||||||||||||||||
| Wandb | $0.03 | $0.13 | 1 | |||||||||||||
Standardwandb/openai/gpt-oss-20b
| ||||||||||||||||
| Novita | $0.04 | $0.15 | 1 | |||||||||||||
Standardnovita/openai/gpt-oss-20b
| ||||||||||||||||
| Ovhcloud | $0.04 | $0.15 | 1 | |||||||||||||
Standardovhcloud/gpt-oss-20b
| ||||||||||||||||
| Together AI | $0.05 | $0.20 | 1 | |||||||||||||
Standardtogether_ai/openai/gpt-oss-20b
| ||||||||||||||||
| Cloudflare is dearest at $0.20 / $0.30 | ||||||||||||||||
Price history
listed seller prices, input and output per 1M tokens Input Output
Input down 71% since first tracked
Cost calculator
estimate your monthly spend on this model$9
estimated / month
Model IDs
copy the exact identifier for your platformopenai/gpt-oss-20b:freedarkbloom/gpt-oss-20bopenai/gpt-oss-20bopenai/gpt-oss-20b:batchwandb/openai/gpt-oss-20bdeepinfra/openai/gpt-oss-20bopenai/gpt-oss-20bovhcloud/gpt-oss-20bnovita/openai/gpt-oss-20btogether_ai/openai/gpt-oss-20bfireworks_ai/accounts/fireworks/models/gpt-oss-20btensormesh/openai/gpt-oss-20bdatabricks/databricks-gpt-oss-20bbedrock_mantle/openai.gpt-oss-20bgroq/openai/gpt-oss-20breplicate/openai/gpt-oss-20bcloudflare/@cf/openai/gpt-oss-20b
Frequently asked questions
gpt-oss-20b pricing, context and availabilityIs gpt-oss-20b still available?
gpt-oss-20b was deprecated on 2026-09-15, but 15 serving providers still list it.
How much does gpt-oss-20b cost?
gpt-oss-20b costs $0.0145 per 1M input tokens from Darkbloom, and $0.07 per 1M output tokens from Darkbloom. Across 15 serving providers, input prices range from $0.0145 to $0.20 per 1M tokens.
What is the context window of gpt-oss-20b?
gpt-oss-20b accepts up to 131K tokens of context and can return up to 131K output tokens.
Which providers serve gpt-oss-20b?
gpt-oss-20b is available from 15 serving providers, each with its own pricing and model ID. Darkbloom is currently the cheapest.
What can gpt-oss-20b do?
gpt-oss-20b supports tool use, reasoning, prompt caching, structured output and web search.
Other OpenAI models
compare pricing across the OpenAI lineupPopular comparisons
head-to-head pages featuring gpt-oss-20b Listed prices from LiteLLM, OpenRouter and Vercel · last checked Oct 14, 2026
Every weekday
AI moves fast. Here's your debrief.
News, analysis, tools, and more.
For people who build with AI