GPT-OSS 20B
openai/gpt-oss-20bHugging FaceChat8-bitconversationaleval-resultsgpt_ossmxfp4reasoningtext-generationtoolsvllm
Providers
Live endpoints for this model, cheapest first. TTFT is time to first token, measured on streamed requests; response is request to last byte, measured on every request; throughput covers the last day; uptime is measured over the last day and week.
| Provider | Context | Max out | Input | Output | Cache read | TTFT | Response | Throughput | Capacity | Uptime 1d | Uptime 7d | Tools | Reasoning | Health | Data |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Consensus Protocolmxfp490% offUS | 131K | 33K | $0.004 | $0.019 | $0.001 | 279ms30m | 2.40s30m | 87 tok/s | 32/32 free | 100.0% | 94.5% | Verified | Effort | Healthy | ZDR |
Pricing
USD per million tokens.
| Provider | Input | Output | Cache read |
|---|---|---|---|
| Consensus Protocolmxfp490% off | $0.004 | $0.019 | $0.001 |
Price history
Input price · USD per million tokens
Output price · USD per million tokens
Performance
Timing percentiles cover the labelled window: the last 30 minutes when it has enough samples, otherwise the last day. TTFT is the seconds to the first token, measured only on streamed requests; response is the seconds from request to last byte, measured on every request, streamed or not. Throughput is output tokens per second over the last day, counted only over streamed generations long enough to have a rate — a buffered answer has no generation phase, and a two-token one times the scheduler rather than the model. Tool calls is the share of the last day of tool requests this endpoint answered with a tool call, published once the day holds twenty of them; under that the cell is blank rather than a rate nothing measured.
| Provider | TTFT p50 | p75 | p90 | p99 | Response p50 | p75 | p90 | p99 | Throughput p50 | p75 | p90 | p99 | Tool calls 1d | Runaways 1d |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Consensus Protocolmxfp4 | 279ms30m | 352ms | 1.13s | 3.21s | 2.40s30m | 11.0s | 37.2s | 196.1s | 87 tok/s | 132 tok/s | 165 tok/s | 198 tok/s | — | 1.5% |
Time to first token by hour · last 7 days
Hourly median per provider. The line spans hours with no streamed traffic; where samples are sparse, a dot marks each hour actually timed.
Throughput by hour · last 7 days
Hourly median output tokens per second per provider, over generations long enough to have a rate. The line spans hours with no answer long enough to measure; where samples are sparse, a dot marks each hour actually measured.
Rotating benchmark probes
Diagnostic scores from rotating dataset subsets. Percentages cover graded replies only and are not full benchmark results. Each score belongs to the provider’s serving configuration.
| Provider | GSM8K | MMLU | aime25 | math500 | mmlu-pro |
|---|---|---|---|---|---|
| Consensus Protocolmxfp4 | 97.8%730 graded | 87.8%984 graded | 81.3%91 graded | 98.1%678 graded | 75.8%719 graded |
Benchmark answers are graded mechanically against the vendored dataset reference answers — numeric comparison of the final number for GSM8K, MATH-500 and AIME 2025, the committed multiple-choice letter for MMLU, MMLU-Pro, MMMU and MMBench, cosine ranking of the returned vectors for the Banking77 and SciDocs retrieval triplets, and per-utterance word match for LibriSpeech and FLEURS — never by another model; a reply that ran into the probe’s token cap before it committed to an answer is recorded but not graded.
Uptime
Share of successful requests and probes over the last day and week.
- Consensus Protocolmxfp41d 100.0%7d 94.5%Healthy
Availability by hour · last 7 days
Share of successful requests and probes per hour and provider. The line spans hours with nothing to measure; where samples are sparse, a dot marks each hour actually measured.
Activity
Prompt tokens and completion tokens per UTC day over the last 30 days. Every billed request counts, the platform's own paid probe traffic included.
curl https://api-staging.heimdallm.com/api/v1/chat/completions \ -H "Authorization: Bearer $HEIMDALL_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "openai/gpt-oss-20b", "messages": [{ "role": "user", "content": "Hello" }] }'Set HEIMDALL_API_KEY to a key from your account. See the quickstart for routing suffixes and provider preferences.
About
Welcome to the gpt-oss series, OpenAI’s open-weight models designed for powerful reasoning, agentic tasks, and versatile developer use cases.
- Author
- openai
- Family
- openai/gpt-oss-20b
- Weights
- openai/gpt-oss-20b
- Parameters
- frequency_penalty (-2–2), logit_bias (any value), logprobs (true/false), max_tokens (1–32768), min_p (0–1), presence_penalty (-2–2), reasoning_effort (low, medium, high), response_format (any value), seed (0–2147483647), stop (list), temperature (0–2), tool_choice (any value), tools (true/false), top_k (0–200), top_p (0–1)
- Licence
- apache-2.0
- Size
- 20.9B parameters, U8 weights
- Tokenizer
- —
- Available until
- —
From the model card
What the repository states about these weights, read from openai/gpt-oss-20b on Oct 11, 2026.
- Languages
- —
- Papers
- 2508.10925
- Datasets
- —
- Revision
- 6cee5e8Aug 26, 2025
- Downloads
- 6.7M
- Likes
- 5.1K