Qwen: Qwen3 235B A22B Instruct 2507 provider comparison

qwen/qwen3-235b-a22b-2507

Qwen3-235B-A22B-Instruct-2507 is a multilingual, instruction-tuned mixture-of-experts language model based on the Qwen3-235B architecture, with 22B active parameters per forward pass. It is optimized for general-purpose text generation, including instruction following,...

Endpoints:: 12
Context:: 262.1K
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

NovitaAInovita/fp8

$0.2533 / 1M tokens

21.53 s estimated response

Cheapest

DeepInfradeepinfra/fp8

$0.2433 / 1M tokens

$0.000365 for the sample request

Fastest

Alibaba Cloud Int.alibaba

14.00 s estimated response

$0.299 / 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.

DeepInfra has the lowest estimated cost for 1,000 input and 500 output tokens at $0.000365.

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. NovitaAI 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 Qwen: Qwen3 235B A22B Instruct 2507
Provider / route	Current pricing ↗	Speed measurements	Endpoint details
Provider / route DeepInfra`deepinfra/fp8` Cheapest	Current pricing Input price $0.09 / 1M tokens Output price $0.55 / 1M tokens Sample cost $0.000365	Speed measurements 500-token response 25.35 s Response starts 0.35 s Output speed 20.0 tok/s	Endpoint details Uptime (30 min) 99.72% Context 262.1K Model format fp8 Max output 16.4K 15 parameters frequency_penalty, logit_bias, max_tokens, min_p, presence_penalty, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_p
Provider / route NovitaAI`novita/fp8` Best cost/speed trade-off	Current pricing Input price $0.09 / 1M tokens Output price $0.58 / 1M tokens Sample cost $0.00038	Speed measurements 500-token response 21.53 s Response starts 0.70 s Output speed 24.0 tok/s	Endpoint details Uptime (30 min) 99.46% Context 131.1K Model format fp8 Max output 16.4K 13 parameters frequency_penalty, max_tokens, presence_penalty, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_p
Provider / route Parasail`parasail/fp8`	Current pricing Input price $0.14 / 1M tokens Output price $0.8 / 1M tokens Sample cost $0.00054	Speed measurements 500-token response 17.67 s Response starts 0.43 s Output speed 29.0 tok/s	Endpoint details Uptime (30 min) 99.94% Context 131.1K Model format fp8 Max output 131.1K 16 parameters frequency_penalty, logit_bias, logprobs, max_tokens, presence_penalty, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p
Provider / route Alibaba Cloud Int.`alibaba` Fastest	Current pricing Input price $0.1495 / 1M tokens Output price $0.598 / 1M tokens Sample cost $0.0004485 Discount 35.0%	Speed measurements 500-token response 14.00 s Response starts 0.49 s Output speed 37.0 tok/s	Endpoint details Uptime (30 min) 99.98% Context 131.1K Model format unknown Max output 32.8K 10 parameters logprobs, max_tokens, presence_penalty, response_format, seed, temperature, tool_choice, tools, top_logprobs, top_p
Provider / route Venice`venice/fp8`	Current pricing Input price $0.15 / 1M tokens Output price $0.75 / 1M tokens Sample cost $0.000525	Speed measurements 500-token response 45.95 s Response starts 0.50 s Output speed 11.0 tok/s	Endpoint details Uptime (30 min) 99.08% Context 128K Model format fp8 Max output 16.4K 11 parameters frequency_penalty, max_tokens, presence_penalty, response_format, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_p
Provider / route Nebius`nebius/fp8`	Current pricing Input price $0.2 / 1M tokens Output price $0.6 / 1M tokens Sample cost $0.0005	Speed measurements 500-token response 15.56 s Response starts 0.41 s Output speed 33.0 tok/s	Endpoint details Uptime (30 min) 96.94% Context 262.1K Model format fp8 Max output N/a 14 parameters frequency_penalty, logit_bias, max_tokens, presence_penalty, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_p
Provider / route Friendli`friendli`	Current pricing Input price $0.2 / 1M tokens Output price $0.8 / 1M tokens Sample cost $0.0006	Speed measurements 500-token response 20.35 s Response starts 0.35 s Output speed 25.0 tok/s	Endpoint details Uptime (30 min) 99.87% Context 262.1K Model format unknown Max output 262.1K 15 parameters frequency_penalty, logprobs, max_tokens, min_p, presence_penalty, repetition_penalty, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p
Provider / route AtlasCloud`atlas-cloud/fp8`	Current pricing Input price $0.2 / 1M tokens Output price $0.88 / 1M tokens Sample cost $0.00064	Speed measurements 500-token response 167.85 s Response starts 1.19 s Output speed 3.0 tok/s	Endpoint details Uptime (30 min) 100.00% Context 131.1K Model format fp8 Max output 131.1K 17 parameters frequency_penalty, logit_bias, logprobs, max_tokens, min_p, presence_penalty, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p
Provider / route StreamLake`streamlake`	Current pricing Input price $0.21 / 1M tokens Output price $0.84 / 1M tokens Sample cost $0.00063 Discount 40.0%	Speed measurements 500-token response 30.05 s Response starts 0.64 s Output speed 17.0 tok/s	Endpoint details Uptime (30 min) 99.87% Context 128K Model format unknown Max output 32K 12 parameters frequency_penalty, logprobs, max_tokens, presence_penalty, response_format, stop, structured_outputs, temperature, tool_choice, tools, top_logprobs, top_p
Provider / route Google Vertex`google-vertex/us-south1`	Current pricing Input price $0.22 / 1M tokens Output price $0.88 / 1M tokens Sample cost $0.00066	Speed measurements 500-token response 15.20 s Response starts 0.50 s Output speed 34.0 tok/s	Endpoint details Uptime (30 min) 100.00% Context 262.1K Model format unknown Max output 16.4K 15 parameters frequency_penalty, logprobs, max_tokens, presence_penalty, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p
Provider / route Google Vertex`google-vertex/us-south1`	Current pricing Input price $0.25 / 1M tokens Output price $1 / 1M tokens Sample cost $0.00075	Speed measurements 500-token response 15.20 s Response starts 0.50 s Output speed 34.0 tok/s	Endpoint details Uptime (30 min) 99.64% Context 262.1K Model format unknown Max output 16.4K 5 parameters max_tokens, response_format, seed, temperature, top_p
Provider / route GMICloud`gmicloud/fp8`	Current pricing Input price $0.35 / 1M tokens Output price $1.4 / 1M tokens Sample cost $0.00105	Speed measurements 500-token response 251.02 s Response starts 1.02 s Output speed 2.0 tok/s	Endpoint details Uptime (30 min) 99.96% Context 262.1K Model format fp8 Max output N/a 10 parameters logprobs, max_tokens, response_format, seed, structured_outputs, temperature, tool_choice, tools, top_logprobs, top_p

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

Qwen3 235B A22B Instruct 2507 deployment guide

How to choose a Qwen3 235B A22B Instruct 2507 inference provider

Qwen3 235B A22B Instruct 2507 is indexed here as a text model by Qwen, with 12 provider endpoints from DeepInfra, NovitaAI, Parasail, Alibaba Cloud Int., Venice, 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 262.1K-token context window. It accepts text input and returns text output. 17 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

DeepInfra currently has the lowest estimated cost for the standard 1,000-input/500-output-token sample at $0.000365. 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

Alibaba Cloud Int. currently has the shortest estimated 500-token response at 14.00 seconds. NovitaAI is the single endpoint closest to the current ideal cost/speed combination. Recent observations can change, so validate finalists with your own prompts.