OpenAI: gpt-oss-120b provider comparison

openai/gpt-oss-120b

gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI designed for high-reasoning, agentic, and general-purpose production use cases. It activates 5.1B parameters per forward pass and is optimized...

Endpoints:: 21
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

Google Vertexgoogle-vertex/global

$0.18 / 1M tokens

3.62 s estimated response

Cheapest

OpenInferenceopen-inference/int8

$0.07 / 1M tokens

$0.000105 for the sample request

Fastest

Cerebrascerebras/fp16

0.86 s estimated response

$0.4833 / 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.

OpenInference has the lowest estimated cost for 1,000 input and 500 output tokens at $0.000105.

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.

20 providers have complete pricing and speed data. Google Vertex 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-120b
Provider / route	Current pricing ↗	Speed measurements	Endpoint details
Provider / route OpenInference`open-inference/int8` Cheapest	Current pricing Input price $0.03 / 1M tokens Output price $0.15 / 1M tokens Sample cost $0.000105	Speed measurements 500-token response 13.60 s Response starts 0.45 s Output speed 38.0 tok/s	Endpoint details Uptime (30 min) 99.35% Context 131.1K Model format int8 Max output 131.1K 13 parameters include_reasoning, max_tokens, min_p, reasoning, reasoning_effort, seed, stop, temperature, tool_choice, tools, top_a, top_k, top_p
Provider / route DekaLLM`dekallm/bf16`	Current pricing Input price $0.03 / 1M tokens Output price $0.18 / 1M tokens Sample cost $0.00012	Speed measurements 500-token response 17.72 s Response starts 1.05 s Output speed 30.0 tok/s	Endpoint details Uptime (30 min) 96.36% Context 131.1K Model format bf16 Max output N/a 17 parameters frequency_penalty, include_reasoning, logit_bias, logprobs, max_tokens, presence_penalty, reasoning, reasoning_effort, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_logprobs, top_p
Provider / route DeepInfra`deepinfra/bf16`	Current pricing Input price $0.037 / 1M tokens Output price $0.17 / 1M tokens Sample cost $0.000122	Speed measurements 500-token response 20.06 s Response starts 7.24 s Output speed 39.0 tok/s	Endpoint details Uptime (30 min) 73.10% 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 Weights & Biases`wandb/fp4`	Current pricing Input price $0.04 / 1M tokens Output price $0.14 / 1M tokens Sample cost $0.00011	Speed measurements 500-token response 13.67 s Response starts 0.52 s Output speed 38.0 tok/s	Endpoint details Uptime (30 min) 97.90% 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 Mancer`mancer/fp4`	Current pricing Input price $0.045 / 1M tokens Output price $0.375 / 1M tokens Sample cost $0.0002325	Speed measurements 500-token response 14.36 s Response starts 1.21 s Output speed 38.0 tok/s	Endpoint details Uptime (30 min) 91.81% Context 131.1K Model format fp4 Max output 131.1K 15 parameters frequency_penalty, include_reasoning, logit_bias, max_tokens, min_p, presence_penalty, reasoning, reasoning_effort, repetition_penalty, seed, stop, temperature, top_a, top_k, top_p
Provider / route NovitaAI`novita/fp4`	Current pricing Input price $0.05 / 1M tokens Output price $0.25 / 1M tokens Sample cost $0.000175	Speed measurements 500-token response 12.65 s Response starts 0.75 s Output speed 42.0 tok/s	Endpoint details Uptime (30 min) 83.57% Context 131.1K Model format fp4 Max output 32.8K 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 SiliconFlow`siliconflow/fp8`	Current pricing Input price $0.05 / 1M tokens Output price $0.45 / 1M tokens Sample cost $0.000275	Speed measurements 500-token response 20.53 s Response starts 1.30 s Output speed 26.0 tok/s	Endpoint details Uptime (30 min) 88.16% 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 DigitalOcean`digitalocean`	Current pricing Input price $0.07 / 1M tokens Output price $0.49 / 1M tokens Sample cost $0.000315	Speed measurements 500-token response 24.38 s Response starts 2.64 s Output speed 23.0 tok/s	Endpoint details Uptime (30 min) 99.30% Context 128K Model format unknown Max output N/a 8 parameters include_reasoning, logprobs, max_tokens, reasoning, reasoning_effort, temperature, top_logprobs, top_p
Provider / route Google Vertex`google-vertex/global` Best cost/speed trade-off	Current pricing Input price $0.09 / 1M tokens Output price $0.36 / 1M tokens Sample cost $0.00027	Speed measurements 500-token response 3.62 s Response starts 0.37 s Output speed 154.0 tok/s	Endpoint details Uptime (30 min) 98.29% Context 131.1K Model format unknown Max output 131.1K 11 parameters include_reasoning, max_tokens, reasoning, reasoning_effort, repetition_penalty, response_format, seed, structured_outputs, temperature, top_k, top_p
Provider / route Baseten`baseten/fp4`	Current pricing Input price $0.1 / 1M tokens Output price $0.5 / 1M tokens Sample cost $0.00035	Speed measurements 500-token response 2.43 s Response starts 0.25 s Output speed 229.0 tok/s	Endpoint details Uptime (30 min) 99.91% Context 128.1K Model format fp4 Max output 128.1K 13 parameters frequency_penalty, include_reasoning, max_tokens, presence_penalty, reasoning, reasoning_effort, response_format, stop, structured_outputs, temperature, tool_choice, tools, top_p
Provider / route Parasail`parasail/fp4`	Current pricing Input price $0.1 / 1M tokens Output price $0.75 / 1M tokens Sample cost $0.000475	Speed measurements 500-token response 5.67 s Response starts 0.41 s Output speed 95.0 tok/s	Endpoint details Uptime (30 min) 95.21% 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 SambaNova`sambanova`	Current pricing Input price $0.14 / 1M tokens Output price $0.95 / 1M tokens Sample cost $0.000615	Speed measurements 500-token response 3.61 s Response starts 0.58 s Output speed 165.0 tok/s	Endpoint details Uptime (30 min) 99.98% Context 131.1K Model format unknown Max output 131.1K 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 $0.15 / 1M tokens Output price $0.6 / 1M tokens Sample cost $0.00045	Speed measurements 500-token response 3.15 s Response starts 1.28 s Output speed 267.0 tok/s	Endpoint details Uptime (30 min) 100.00% Context 131.1K Model format unknown Max output N/a 8 parameters include_reasoning, max_tokens, reasoning, reasoning_effort, stop, temperature, top_k, top_p
Provider / route DeepInfra`deepinfra/turbo`	Current pricing Input price $0.15 / 1M tokens Output price $0.6 / 1M tokens Sample cost $0.00045	Speed measurements 500-token response 6.86 s Response starts 0.97 s Output speed 85.0 tok/s	Endpoint details Uptime (30 min) 99.97% Context 131.1K Model format bf16 Max output 16.4K 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 Together`together`	Current pricing Input price $0.15 / 1M tokens Output price $0.6 / 1M tokens Sample cost $0.00045	Speed measurements 500-token response 4.17 s Response starts 0.23 s Output speed 127.0 tok/s	Endpoint details Uptime (30 min) 99.29% Context 131.1K Model format unknown Max output N/a 17 parameters frequency_penalty, include_reasoning, logit_bias, max_tokens, min_p, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_p
Provider / route Nebius`nebius/fp4`	Current pricing Input price $0.15 / 1M tokens Output price $0.6 / 1M tokens Sample cost $0.00045	Speed measurements 500-token response 5.01 s Response starts 0.38 s Output speed 108.0 tok/s	Endpoint details Uptime (30 min) 99.52% Context 131.1K Model format fp4 Max output N/a 17 parameters frequency_penalty, include_reasoning, logit_bias, max_tokens, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_p
Provider / route Amazon Bedrock`amazon-bedrock/eu-west-1`	Current pricing Input price $0.15 / 1M tokens Output price $0.6 / 1M tokens Sample cost $0.00045	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 8 parameters include_reasoning, max_tokens, reasoning, reasoning_effort, stop, temperature, top_k, top_p
Provider / route Phala`phala`	Current pricing Input price $0.15 / 1M tokens Output price $0.6 / 1M tokens Sample cost $0.00045	Speed measurements 500-token response 11.07 s Response starts 1.07 s Output speed 50.0 tok/s	Endpoint details Uptime (30 min) 95.20% Context 131.1K Model format unknown Max output 131.1K 19 parameters frequency_penalty, include_reasoning, 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
Provider / route Groq`groq`	Current pricing Input price $0.15 / 1M tokens Output price $0.6 / 1M tokens Sample cost $0.00045	Speed measurements 500-token response 1.47 s Response starts 0.15 s Output speed 379.0 tok/s	Endpoint details Uptime (30 min) 100.00% 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
Provider / route MARA`mara`	Current pricing Input price $0.15 / 1M tokens Output price $0.75 / 1M tokens Sample cost $0.000525	Speed measurements 500-token response 3.53 s Response starts 1.01 s Output speed 198.0 tok/s	Endpoint details Uptime (30 min) 84.54% Context 131.1K Model format unknown Max output 131.1K 14 parameters include_reasoning, logprobs, max_tokens, reasoning, reasoning_effort, response_format, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p
Provider / route Cerebras`cerebras/fp16` Fastest	Current pricing Input price $0.35 / 1M tokens Output price $0.75 / 1M tokens Sample cost $0.000725	Speed measurements 500-token response 0.86 s Response starts 0.25 s Output speed 829.5 tok/s	Endpoint details Uptime (30 min) 99.92% Context 131.1K Model format fp16 Max output 41K 17 parameters frequency_penalty, include_reasoning, logit_bias, logprobs, max_tokens, presence_penalty, reasoning, reasoning_effort, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_logprobs, top_p

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

gpt-oss-120b deployment guide

How to choose a gpt-oss-120b inference provider

gpt-oss-120b is indexed here as a text model by OpenAI, with 21 provider endpoints from OpenInference, DekaLLM, DeepInfra, Weights & Biases, Mancer, 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. 21 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

OpenInference currently has the lowest estimated cost for the standard 1,000-input/500-output-token sample at $0.000105. 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

Cerebras currently has the shortest estimated 500-token response at 0.86 seconds. Google Vertex is the single endpoint closest to the current ideal cost/speed combination. Recent observations can change, so validate finalists with your own prompts.