OpenAI: gpt-oss-20b provider comparison

openai/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...

Endpoints:: 13
Context:: 131.1K
Input:: text
Output:: text

Compare this model in a set →

At a glance

Comparison winners

Based on 1,000 input and 500 output tokens using the latest published median speed data.

Best cost/speed trade-off

Weights & Biaseswandb/fp4

$0.0633 / 1M tokens

3.87 s estimated response

Cheapest

Weights & Biaseswandb/fp4

$0.0633 / 1M tokens

$0.000095 for the sample request

Fastest

Groqgroq

1.21 s estimated response

$0.15 / 1M tokens

Visual comparison

Price and performance charts

Compare published endpoint pricing and estimated response time visually.

Effective price by endpoint

USD per 1 million tokens for input and output. Lower is better.

Weights & Biases has the lowest estimated cost for 1,000 input and 500 output tokens at $0.000095.

Estimated cost vs. response time

Based on 1,000 input and 500 output tokens. Cost is shown in USD per 1 million tokens. Lower and further left is better; the single best cost/speed trade-off appears at full opacity.

12 providers have complete pricing and speed data. Weights & Biases is closest to the ideal combination of lowest cost and fastest response.

Server-rendered comparison

Provider endpoints

Prompt and completion prices are effective USD per token as published by OpenRouter. Missing speed data never removes an endpoint.

Rank providers by

Endpoint comparison for OpenAI: gpt-oss-20b
Provider / route	Current pricing ↗	Speed measurements	Endpoint details
Provider / route DekaLLM`dekallm/bf16`	Current pricing Input price $0.029 / 1M tokens Output price $0.14 / 1M tokens Sample cost $0.000099	Speed measurements 500-token response 30.36 s Response starts 0.95 s Output speed 17.0 tok/s	Endpoint details Uptime (30 min) 96.20% Context 131.1K Model format bf16 Max output N/a 15 parameters frequency_penalty, include_reasoning, logit_bias, logprobs, max_tokens, presence_penalty, reasoning, reasoning_effort, response_format, seed, stop, structured_outputs, temperature, top_logprobs, top_p
Provider / route Weights & Biases`wandb/fp4` CheapestBest cost/speed trade-off	Current pricing Input price $0.03 / 1M tokens Output price $0.13 / 1M tokens Sample cost $0.000095	Speed measurements 500-token response 3.87 s Response starts 0.27 s Output speed 139.0 tok/s	Endpoint details Uptime (30 min) 99.87% Context 131.1K Model format fp4 Max output 131.1K 18 parameters frequency_penalty, include_reasoning, logprobs, max_tokens, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p
Provider / route DeepInfra`deepinfra/bf16`	Current pricing Input price $0.03 / 1M tokens Output price $0.14 / 1M tokens Sample cost $0.0001	Speed measurements 500-token response 5.65 s Response starts 0.74 s Output speed 102.0 tok/s	Endpoint details Uptime (30 min) 99.99% Context 131.1K Model format bf16 Max output 131.1K 18 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
Provider / route NovitaAI`novita/fp4`	Current pricing Input price $0.04 / 1M tokens Output price $0.15 / 1M tokens Sample cost $0.000115	Speed measurements 500-token response 7.58 s Response starts 0.44 s Output speed 70.0 tok/s	Endpoint details Uptime (30 min) 100.00% Context 131.1K Model format fp4 Max output 32.8K 16 parameters frequency_penalty, include_reasoning, logprobs, max_tokens, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, top_k, top_logprobs, top_p
Provider / route Phala`phala`	Current pricing Input price $0.04 / 1M tokens Output price $0.15 / 1M tokens Sample cost $0.000115	Speed measurements 500-token response 10.05 s Response starts 0.61 s Output speed 53.0 tok/s	Endpoint details Uptime (30 min) 100.00% Context 131.1K Model format unknown Max output 131.1K 13 parameters frequency_penalty, include_reasoning, max_tokens, min_p, presence_penalty, reasoning, reasoning_effort, repetition_penalty, seed, stop, temperature, top_k, top_p
Provider / route SiliconFlow`siliconflow/fp8`	Current pricing Input price $0.04 / 1M tokens Output price $0.18 / 1M tokens Sample cost $0.00013	Speed measurements 500-token response 42.98 s Response starts 1.32 s Output speed 12.0 tok/s	Endpoint details Uptime (30 min) 100.00% Context 131.1K Model format fp8 Max output 8.2K 10 parameters frequency_penalty, include_reasoning, max_tokens, reasoning, reasoning_effort, response_format, structured_outputs, temperature, top_k, top_p
Provider / route Parasail`parasail/fp4`	Current pricing Input price $0.04 / 1M tokens Output price $0.2 / 1M tokens Sample cost $0.00014	Speed measurements 500-token response 6.78 s Response starts 0.53 s Output speed 80.0 tok/s	Endpoint details Uptime (30 min) 96.50% Context 131.1K Model format fp4 Max output 131.1K 19 parameters frequency_penalty, include_reasoning, logit_bias, logprobs, max_tokens, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p
Provider / route Together`together`	Current pricing Input price $0.05 / 1M tokens Output price $0.2 / 1M tokens Sample cost $0.00015	Speed measurements 500-token response 4.43 s Response starts 0.23 s Output speed 119.0 tok/s	Endpoint details Uptime (30 min) Insufficient recent data Context 131.1K Model format unknown Max output N/a 15 parameters frequency_penalty, include_reasoning, logit_bias, max_tokens, min_p, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, stop, structured_outputs, temperature, top_k, top_p
Provider / route Amazon Bedrock`amazon-bedrock/eu-west-1`	Current pricing Input price $0.07 / 1M tokens Output price $0.15 / 1M tokens Sample cost $0.000145	Speed measurements 500-token response Insufficient recent data Response starts Insufficient recent data Output speed Insufficient recent data	Endpoint details Uptime (30 min) Insufficient recent data Context 131.1K Model format unknown Max output N/a 10 parameters include_reasoning, max_tokens, reasoning, reasoning_effort, stop, temperature, tool_choice, tools, top_k, top_p
Provider / route Amazon Bedrock`amazon-bedrock`	Current pricing Input price $0.07 / 1M tokens Output price $0.15 / 1M tokens Sample cost $0.000145	Speed measurements 500-token response 15.17 s Response starts 7.23 s Output speed 63.0 tok/s	Endpoint details Uptime (30 min) 96.93% Context 131.1K Model format unknown Max output N/a 10 parameters include_reasoning, max_tokens, reasoning, reasoning_effort, stop, temperature, tool_choice, tools, top_k, top_p
Provider / route Google Vertex`google-vertex/us-central1`	Current pricing Input price $0.07 / 1M tokens Output price $0.25 / 1M tokens Sample cost $0.000195	Speed measurements 500-token response 4.86 s Response starts 1.69 s Output speed 158.0 tok/s	Endpoint details Uptime (30 min) 89.92% Context 131.1K Model format unknown Max output 32.8K 11 parameters include_reasoning, max_tokens, reasoning, reasoning_effort, repetition_penalty, response_format, seed, structured_outputs, temperature, top_k, top_p
Provider / route Fireworks`fireworks`	Current pricing Input price $0.07 / 1M tokens Output price $0.3 / 1M tokens Sample cost $0.00022	Speed measurements 500-token response 4.10 s Response starts 0.31 s Output speed 132.0 tok/s	Endpoint details Uptime (30 min) 99.66% Context 131.1K Model format unknown Max output N/a 16 parameters frequency_penalty, include_reasoning, logit_bias, max_tokens, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_p
Provider / route Groq`groq` Fastest	Current pricing Input price $0.075 / 1M tokens Output price $0.3 / 1M tokens Sample cost $0.000225	Speed measurements 500-token response 1.21 s Response starts 0.28 s Output speed 538.0 tok/s	Endpoint details Uptime (30 min) 99.77% Context 131.1K Model format unknown Max output 65.5K 12 parameters include_reasoning, max_tokens, reasoning, reasoning_effort, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_p

Endpoint data fetched Jul 15, 2026, 11:17 PM UTC.

gpt-oss-20b deployment guide

How to choose a gpt-oss-20b inference provider

gpt-oss-20b is indexed here as a text model by OpenAI, with 13 provider endpoints from DekaLLM, Weights & Biases, DeepInfra, NovitaAI, Phala, and others. The comparison preserves each exact OpenRouter routing tag so pricing and performance observations can be connected to the route an application would actually request.

Match the endpoint to the workload

The model publishes a 131.1K-token context window. It accepts text input and returns text output. 20 distinct supported parameters appear across the listed routes. Confirm limits on the specific endpoint rather than assuming every host exposes the same configuration.

Compare the real request economics

Weights & Biases currently has the lowest estimated cost for the standard 1,000-input/500-output-token sample at $0.000095. Input-heavy and output-heavy applications can produce a different result, so review both per-million-token prices in the endpoint table.

Balance response start and generation speed

Groq currently has the shortest estimated 500-token response at 1.21 seconds. Weights & Biases is the single endpoint closest to the current ideal cost/speed combination. Recent observations can change, so validate finalists with your own prompts.