OpenAIGeneralReleased 5 August 2025
gpt-oss-20b pricing, context and benchmark
gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license.
131K
Input context
118K
Output context
$0.018
Official input
$0.09
Official output
n/a
AA index
01 · Model Overview
Everything that matters, in one table
Context, pricing, cache, modalities and tooling, from the model registry and Artificial Analysis.
| Model | gpt-oss-20b |
| Model ID | openai/gpt-oss-20b |
| Provider | OpenAI |
| Release date | 5 August 2025 |
| Status | Generally available |
| Input context | 131,072 tokens |
| Output context | 117,964 tokens |
| Official input pricing | $0.018 per 1M tokens |
| Official output pricing | $0.09 per 1M tokens |
| Cache read pricing | $0.03 per 1M tokens |
| Market input pricing | $0.018 per 1M (cheapest route) |
| Market output pricing | $0.03 per 1M (cheapest route) |
| Reasoning tokens | Billed as output at $0.09 per 1M |
| Input modalities | text |
| Output modalities | text |
| Reasoning control | Yes, effort levels exposed |
| Tool support | Functions, tool choice, structured outputs |
| Tokenizer | GPT |
| Provider routes | 11 endpoints |
| Artificial Analysis index | Not published |
| Speed | Not published |
| Knowledge cutoff | Not publicly stated |
Input and output support
| Modality | Input | Output | Billed at |
|---|---|---|---|
| Text | ✓ yes | ✓ yes | $0.018 in |
| Image | no | no | not supported |
| Audio | no | no | not supported |
| Video | no | no | not supported |
| File (PDF) | no | no | not supported |
02 · Overview
What is gpt-oss-20b?
gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license.
It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for...
gpt-oss-20b carries a 131K token window, uses the GPT tokenizer and is served by OpenAI. On Hoodgen it sits in the General tier of the model menu next to its siblings, so you can switch mid conversation without losing the thread.
03 · Release
Release and what changed from the previous version
Released5 August 2025
| Property | o3 Pro | gpt-oss-20b | Change |
|---|---|---|---|
| Released | 2025 | 2025 | Same release window |
| Context | 200K | 131K | Changed |
| Max output | 100K | 118K | Changed |
| Input price | $20 | $0.018 | Cheaper |
| Output price | $80 | $0.09 | Cheaper |
| Input modalities | text, file, image | text | Unchanged |
| AA index | n/a | n/a | Newer benchmark run |
04 · Capabilities
Features and capabilities
◈
Thinking control
Reasoning effort can be dialled per request, so one model covers deep analysis and fast chat.
⚙
Function calling
Tools, tool choice and parallel calls, which is what agent loops need to act on results.
{ }
Structured outputs
Schema constrained JSON with response_format, so no parse and retry loop in your code.
⚡
Prompt caching
Cached input costs $0.03 per million instead of $0.018, so stable prefixes get cheaper.
⌁
Seeds
A seed can be pinned so the same prompt is reproducible across test runs.
05 · Reasoning
Thinking and reasoning capabilities
gpt-oss-20b can think before answering and exposes the effort level per request. Reasoning tokens bill as output at $0.09 per million, and the trace can be returned with include_reasoning.
| Level | First token | Cost per task | Best for |
|---|---|---|---|
| Measured | n/a | n/a | Artificial Analysis measures the default reasoning configuration |
| Low effort | Faster | Lower | Chat, autocomplete, classification |
| High effort | Slower | Higher | Debugging, planning, multi step agents |
07 · Use cases
What can you do with gpt-oss-20b?
Everyday writing
Drafts, rewrites, summaries and replies.
Research
Source material condensed with the key numbers kept.
Chat
General questions where latency and cost both matter.
Tooling inside Hoodgen
Writer, Humanizer, Coding and the rest of the tool suite.
08 · Cost levers
Prompt caching and how to cut the bill
Cached input bills at $0.03 per million instead of $0.018. Keep the system prompt and documents first and the changing question last, so every call hits the cache.
09 · Pricing
gpt-oss-20b pricing
Official pricing
List price
$0.018 / 1M in
$0.09 / 1M out
- $0.03 cached input
- 131K context
- text input
Cheapest route
$0.018 / 1M in
$0.03 / 1M out
- 11 routes on OpenRouter
- What Hoodgen routes to first
- Same model, lower bill
Market pricing, every route
| Provider | Tier | Input | Output | Cache read | Uptime 24h |
|---|---|---|---|---|---|
| Darkbloom | standard | $0.018 | $0.09 | $0.009 | 100.0% |
| AkashML | standard | $0.018 | $0.09 | $0.009 | 100.0% |
| PrimeIntellect | standard | $0.02 | $0.03 | $0.018 | 100.0% |
| DekaLLM | standard | $0.029 | $0.14 | $0.029 | 100.0% |
| CoreWeave | standard | $0.03 | $0.13 | $0.03 | 100.0% |
| DeepInfra | standard | $0.03 | $0.14 | n/a | 100.0% |
| Parasail | standard | $0.03 | $0.15 | $0.02 | 100.0% |
| SiliconFlow | standard | $0.04 | $0.18 | n/a | 100.0% |
| Amazon Bedrock | standard | $0.07 | $0.15 | n/a | 100.0% |
| Amazon Bedrock | standard | $0.07 | $0.15 | n/a | 100.0% |
| Groq | standard | $0.075 | $0.3 | $0.037 | 100.0% |
API pricing and cost calculator
| Rate | Per call | Total | Credits |
|---|---|---|---|
| Cheapest route | $0.0003 | $0.15 | 142 |
| List price | $0.0006 | $0.32 | 304 |
10 · Developers
API and developer information
| Model ID | openai/gpt-oss-20b |
| Endpoint | https://openrouter.ai/api/v1/chat/completions |
| Parameters | frequency_penalty · include_reasoning · logit_bias · logprobs · max_tokens · min_p · presence_penalty · reasoning · reasoning_effort · repetition_penalty · response_format · seed · stop · structured_outputs · temperature · tool_choice · tools · top_k · top_logprobs · top_p |
| Streaming | Yes, server sent events |
| Modality string | text->text |
| Max prompt tokens | 131,072 |
| Max completion tokens | 117,964 |
curl https://openrouter.ai/api/v1/chat/completions \
-H "Authorization: Bearer $OPENROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "openai/gpt-oss-20b",
"messages": [{ "role": "user", "content": "Summarise this document." }],
"max_tokens": 1024
}'11 · Limitations
Model limitations
| Text only output | It cannot generate images, audio or video. Pair it with an image model on Hoodgen when you need visuals. |
| Output cap | 117,964 tokens per request. Long deliverables have to be chunked. |
| Benchmarks not published | Artificial Analysis does not publish an Intelligence Index for this release, so no benchmark score is claimed here. |
| Knowledge cutoff | The provider does not state a fixed cutoff for every release. Verify very recent events against search. |
12 · Comparison
gpt-oss-20b vs other OpenAI models
| Model | Context | Input /1M | Output /1M | AA index | Released |
|---|---|---|---|---|---|
| gpt-oss-20b | 131K | $0.018 | $0.09 | n/a | 2025 |
| GPT-5.5 Pro | 1.05M | $30 | $180 | n/a | 2026 |
| GPT-5.4 Pro | 1.05M | $30 | $180 | n/a | 2026 |
| GPT-5.2 Pro | 400K | $21 | $168 | n/a | 2025 |
| o3 Pro | 200K | $20 | $80 | n/a | 2025 |
| GPT-5 Pro | 400K | $15 | $120 | n/a | 2025 |
13 · Privacy
Privacy and data policy
- On Hoodgen, prompts and attachments answer that request and stay in your conversation history so you can reopen it. They are not used to train models.
- Requests route through OpenRouter to OpenAI and are covered by that provider's API terms.
- Do not paste secrets or personal data you would not send to any hosted model.
14 · FAQ
gpt-oss-20b questions, answered
gpt-oss-20b is a general model from OpenAI, released 5 August 2025. gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license.
✦
Start chatting with gpt-oss-20b
Free to try on Hoodgen AI, with a live credit estimate before every send. No API key, no setup.