← Compare all APIs

gpt-oss-20b API prices

Compare routes under the same recorded conditions.

Official pricing source ↗

Channel choices

API: openai/gpt-oss-20b:batch · USD · Batch · bf16 · 131K · Global

Reference: 1M input + 1M output tokens; no cache or extras. Not a bill.

Only 1 channel in this scope. Configuration is incomplete; equivalent replacement and savings are unverified.

View recorded configuration conditions

Context limit: 131072

Input/output limits: Unrecorded / 117964

Regions: global

Billing scope: Unrecorded

Parameter declarations: frequency_penalty, include_reasoning, logit_bias, max_tokens, min_p, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_p

Native tool declarations: Source lists no entries

Implicit caching declaration: No

Account restriction: Unrecorded

Subroute label: No extra label

These are recorded source fields; actual capability, performance and account access remain unverified.

Change billing and configuration scope

DeepInfra

deepinfra/bf16

Purchased through OpenRouter

$0.136

1M input + 1M output token subtotal

Input / output per 1M tokens: $0.024 / $0.112

Lowest reference rate in this group; verify output and access.

Use this endpoint
Official API docs ↗

Create a provider API key and set the local environment variable OPENROUTER_API_KEY, then run the example below.

After submitting, use the official docs to check job status and retrieve results.

Request URL
https://openrouter.ai/api/v1/batches
Connection settings
Base URL: https://openrouter.ai/api/v1
Catalog model: openai/gpt-oss-20b:batch
Request API model: openai/gpt-oss-20b
Protocol: openai
Provider routing: deepinfra/bf16
API key env: OPENROUTER_API_KEY
Docs: https://openrouter.ai/docs/batch-quickstart
curl
curl "https://openrouter.ai/api/v1/batches" -H "Authorization: Bearer $OPENROUTER_API_KEY" -H "Content-Type: application/json" -d '{
  "endpoint": "/v1/chat/completions",
  "model": "openai/gpt-oss-20b",
  "provider": {
    "only": [
      "deepinfra/bf16"
    ],
    "allow_fallbacks": false
  },
  "completion_window": "24h",
  "requests": [
    {
      "custom_id": "example-1",
      "body": {
        "messages": [
          {
            "role": "user",
            "content": "Read this support ticket and return a JSON object with \"summary\", \"urgency\" (low, medium, or high), and \"next_step\". Use null for missing details.\n\nTicket: I paid for the Pro plan yesterday, but my account still shows Free. My team needs access before tomorrow’s demo."
          }
        ]
      }
    }
  ]
}'
Prompt for your coding agent
Use the OpenAI-compatible API at https://openrouter.ai/api/v1 with model "openai/gpt-oss-20b". Use the asynchronous Batch API to obtain the batch rate. At the batch top level set provider.only to ["deepinfra/bf16"] and allow_fallbacks to false. Check the provider and billed cost in the result. Read the API key from the OPENROUTER_API_KEY environment variable; never hardcode it. Pricing observed on 2026-11-22: $0.024 / 1M input tokens, $0.112 / 1M output tokens.  API integration docs: https://openrouter.ai/docs/batch-quickstart. Billing scope and conditions: Served by DeepInfra through OpenRouter, not a direct-provider contract. Published rates for route deepinfra/bf16; bf16 quantization. Other fees, tier conditions and regional access must be checked before use.. Source: https://openrouter.ai/api/v1/models/openai/gpt-oss-20b%3Abatch/endpoints.
Task test prompt
Read this support ticket and return a JSON object with "summary", "urgency" (low, medium, or high), and "next_step". Use null for missing details.

Ticket: I paid for the Pro plan yesterday, but my account still shows Free. My team needs access before tomorrow’s demo.
Price facts
Input: $0.024 / 1M tokens
Output: $0.112 / 1M tokens
Currency: USD
Regions: global
Unit source: api
Checked: 2026-11-22T09:20:26.019Z
Source: https://openrouter.ai/api/v1/models/openai/gpt-oss-20b%3Abatch/endpoints

Source check: 2026-11-22 · valid no later than 2026-11-23 09:21 UTC

Billing evidence and sources

openai/gpt-oss-20b:batch · deepinfra/bf16

Billing conditions

Served by DeepInfra through OpenRouter, not a direct-provider contract. Published rates for route deepinfra/bf16; bf16 quantization. Other fees, tier conditions and regional access must be checked before use.

Cache read / 1M
Not published
Cache write / 1M
Not published
Per request
Not published
Route
deepinfra/bf16 · bf16

Declared parameters: frequency_penalty, include_reasoning, logit_bias, max_tokens, min_p, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_p

Price source ↗
Compare saved channels →
All quotes, filters and usage tools
History · Billing and conditions

gpt-oss-20b API pricing: official and providers: 10 current published quote records across 2 exact API IDs and 7 providers. Compare API versions, currencies, regions and billing conditions separately; prices do not prove task quality or account access.

Latest recorded source check: 2026-11-22

Recorded changes

Latest 100 recorded price and availability changes
ObservedModel / provider routeChangeInput rateOutput rate

OpenAI: gpt-oss-20b

AkashML · akashml/fp4

Source ↗
Availability

active → unavailable

$0.018 / 1M tokens → $0.018 / 1M tokens$0.09 / 1M tokens → $0.09 / 1M tokens

OpenAI: gpt-oss-20b

SiliconFlow · siliconflow/fp8

Source ↗
Availability

active → unavailable

$0.04 / 1M tokens → $0.04 / 1M tokens$0.18 / 1M tokens → $0.18 / 1M tokens

OpenAI: gpt-oss-20b

Darkbloom · darkbloom/fp8

Source ↗
Decrease$0.018 / 1M tokens → $0.017 / 1M tokens$0.09 / 1M tokens → $0.085 / 1M tokens

OpenAI: gpt-oss-20b

PrimeIntellect · primeintellect/fp4

Source ↗
Increase$0.175 / 1M tokens → $0.20 / 1M tokens$0.262 / 1M tokens → $0.30 / 1M tokens

Showing up to 100 recent events. Prices for unavailable endpoints are retained for reference. Failed source checks do not establish that a price stayed unchanged.

Open a quote’s billing details for cache reads/writes, request fees and endpoint instructions. Batch, free, promotional and off-peak tiers are selected separately. Context-tier, storage, search, tax and platform fees may be excluded.

Find lower-cost alternatives

Find lower-cost alternatives

Choose your existing model to find lower reference costs with matching declared technical constraints. Output quality remains unverified.

Loading verified quotes…

Users’ choices · gpt-oss-20b

Describe your work to find related tasks.

Public scores by use

The same model scores differently by use. Once you pick a use, the site applies that use’s quality gate before comparing cost.

Text

  • Chinese1352664 votes · 2026-11-19
  • Classify12812533 votes · 2026-11-19
  • Extract fields12812533 votes · 2026-11-19
  • General text131710393 votes · 2026-11-19
  • Multi-turn12891774 votes · 2026-11-19
  • Summarize13032099 votes · 2026-11-19
  • Translate & rewrite12995882 votes · 2026-11-19
  • Write12411229 votes · 2026-11-19

Code & tools

  • Code13682111 votes · 2026-11-19

Industry

  • Business & finance13221832 votes · 2026-11-19
  • Legal & government1321678 votes · 2026-11-19
  • Medical & health1351562 votes · 2026-11-19

Scores come from Arena public-preference leaderboards (CC BY 4.0), latest published 2026-11-19. They describe the model, not your task’s success rate or a quantized channel. Arena ↗

View independent evaluation on Artificial Analysis ↗ · external link; we do not copy its scores or prices.

Declared model facts

Declared fields from the OpenRouter model catalog, bound to exact API IDs; not our tests, price approval or account availability.

Input
Text
Output
Text
Context
131K
Reasoning parameter
Supported
Knowledge cutoff
2024-06-30
Open-weights clue
openai/gpt-oss-20b
Per-API facts · 2 versions
  • openai/gpt-oss-20b · text → text · 131K · reasoning
  • openai/gpt-oss-20b:batch · text → text · 131K · reasoning

Source: https://openrouter.ai/api/v1/models, captured 2026-11-21T09:20:20.905Z, raw SHA-256: 19214ba490361282… . Unstated facts stay unknown and are not inferred from names or sibling models.

Classification sources

Source declarations for registered versions; not task quality, current account availability or price approval. Unknown versions do not inherit sibling classifications.

Model variants

Snapshots and routing variants remain distinct. Choose an exact ID in the comparison before deciding; model quality and access conditions need your own review.

More from this maker

OpenAI →