This model is deprecated. We only continue performance benchmarking for the default 10k input token workload. Results for other workloads are historical and no longer updated.

Alibaba has launched a newer model, Qwen3.5 397B A17B. We suggest considering it instead.

For more information, see comparison of Qwen3.5 397B A17B to other models and API provider benchmarks for Qwen3.5 397B A17B.

Qwen3 235B A22B 2507 Instruct logo
Alibaba
•

Open weights model

•

Released July 2025

Qwen3 235B A22B 2507 Instruct API Provider Benchmarking & Analysis

Model Comparison

Analysis of API providers for Qwen3 235B A22B 2507 Instruct across performance metrics including latency (time to first token), output speed (output tokens per second), price and others. API providers benchmarked include Nebius, Novita, DeepInfra (FP8), FriendliAI, Hyperbolic, Amazon Bedrock, Parasail, Scaleway, Alibaba Cloud, Google Vertex.

Fastest

#1
Google VertexGoogle Vertex
100.7 t/s
#2
ScalewayScaleway
86.8 t/s
#3
AmazonAmazon
66.9 t/s
#4
NebiusNebius
63.1 t/s
#5
Alibaba CloudAlibaba Cloud
56.1 t/s

Output speed

Total 10 providers

Lowest Latency

#1
DeepInfra (FP8)DeepInfra (FP8)
0.60 s
#2
Google VertexGoogle Vertex
0.79 s
#3
NebiusNebius
1.36 s
#4
AmazonAmazon
1.37 s
#5
ParasailParasail
1.47 s

Time to first token

Total 10 providers

Lowest Price

#1
DeepInfra (FP8)DeepInfra (FP8)
$0.14
#2
NovitaNovita
$0.14
#3
ParasailParasail
$0.15
#4
NebiusNebius
$0.24
#5
FriendliAIFriendliAI
$0.26

Blended price (per 1M tokens)

Total 10 providers

Qwen3 235B 2507 is available through 10 API providers, each offering different performance characteristics and pricing. Below is a comparison of the key metrics across providers.

  • For output speed, the top providers are Google Vertex (100.7 t/s), Scaleway (86.8 t/s), Amazon (66.9 t/s). Speed varies significantly across providers, with a 506% difference between the fastest and slowest.
  • For latency, DeepInfra (FP8) (0.60s), Google Vertex (0.79s), Nebius (1.36s) offer the lowest time to first token.
  • For pricing, DeepInfra (FP8) (0.14), Novita (0.14), Parasail (0.15) offer the lowest blended prices per 1M tokens. Prices vary up to 14.7x across providers.
  • DeepInfra (FP8) offers both the lowest latency and best pricing, making it attractive for cost-conscious applications. Google Vertex is the fastest option for throughput-intensive workloads.

Highlights

Speed

Output tokens per second · Higher is better

End-to-End Response Time

Seconds · Lower is better

Price

USD per 1M tokens (blended) · Lower is better

Pricing

Pricing: Cache Hit, Input, and Output

Price (USD per M Tokens) · Lower is better · 10,000 input tokens

Price per token for cached prompts (previously processed), typically offering a significant discount compared to regular input price, represented as USD per million tokens. The values shown here are the cache hit price; cache write and cache storage are billed separately and vary by provider — see "Cache pricing by provider" for detail.

Price per token included in the request/message sent to the API, represented as USD per million Tokens.

The blended cache price shown here uses cache hit price only. Other caching costs differ by provider:

  • Anthropic: charges a separate cache write fee, with different rates for 5-minute and 1-hour TTLs (1-hour TTL is more expensive).
  • Google (Vertex/Gemini): charges a per-hour cache storage fee in addition to cache hit pricing. Some providers also use tiered pricing for prompts above 200K tokens.
  • OpenAI, DeepSeek, others: typically charge only cache hit pricing with no write or storage fee.

See Prompt Caching for the full breakdown.

Price per token generated by the model (received from the API), represented as USD per million Tokens.

Pricing: Blended Price

Blended at 7:2:1 (cache-input-output) · USD per 1M tokens (blended) · Lower is better

Price per token, shown in USD per million tokens. Price is a blend of cache hit, input, and output token prices using the selected ratio (default 7:2:1 cache-input-output).

Price per token for cached prompts (previously processed), typically offering a significant discount compared to regular input price, represented as USD per million tokens. The values shown here are the cache hit price; cache write and cache storage are billed separately and vary by provider — see "Cache pricing by provider" for detail.

The blended cache price shown here uses cache hit price only. Other caching costs differ by provider:

  • Anthropic: charges a separate cache write fee, with different rates for 5-minute and 1-hour TTLs (1-hour TTL is more expensive).
  • Google (Vertex/Gemini): charges a per-hour cache storage fee in addition to cache hit pricing. Some providers also use tiered pricing for prompts above 200K tokens.
  • OpenAI, DeepSeek, others: typically charge only cache hit pricing with no write or storage fee.

See Prompt Caching for the full breakdown.

Figures represent median (P50) measurement over the past 72 hours to reflect sustained changes in performance.

Pricing: Cache Discount

1 - (cache hit price / input price) · Higher is better

Reduction in input token cost due to cache hit relative to input price. Formula: 1 - (Cache Hit Price per Token / Input Token Price), where cache hit price is the first-party cache hit price or the median provider cache hit price. Note that this discount figure does not account for all costs associated with cache hits, such as cache write and storage costs.

Price per token for cached prompts (previously processed), typically offering a significant discount compared to regular input price, represented as USD per million tokens. The values shown here are the cache hit price; cache write and cache storage are billed separately and vary by provider — see "Cache pricing by provider" for detail.

The blended cache price shown here uses cache hit price only. Other caching costs differ by provider:

  • Anthropic: charges a separate cache write fee, with different rates for 5-minute and 1-hour TTLs (1-hour TTL is more expensive).
  • Google (Vertex/Gemini): charges a per-hour cache storage fee in addition to cache hit pricing. Some providers also use tiered pricing for prompts above 200K tokens.
  • OpenAI, DeepSeek, others: typically charge only cache hit pricing with no write or storage fee.

See Prompt Caching for the full breakdown.

Figures represent median (P50) measurement over the past 72 hours to reflect sustained changes in performance.

Output Speed vs. Price

Blended at 7:2:1 (cache-input-output) · Output speed: output tokens per second · USD per 1M tokens (blended)
Most attractive quadrant
Pareto line

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

Price per token, shown in USD per million tokens. Price is a blend of cache hit, input, and output token prices using the selected ratio (default 7:2:1 cache-input-output).

Price per token for cached prompts (previously processed), typically offering a significant discount compared to regular input price, represented as USD per million tokens. The values shown here are the cache hit price; cache write and cache storage are billed separately and vary by provider — see "Cache pricing by provider" for detail.

The blended cache price shown here uses cache hit price only. Other caching costs differ by provider:

  • Anthropic: charges a separate cache write fee, with different rates for 5-minute and 1-hour TTLs (1-hour TTL is more expensive).
  • Google (Vertex/Gemini): charges a per-hour cache storage fee in addition to cache hit pricing. Some providers also use tiered pricing for prompts above 200K tokens.
  • OpenAI, DeepSeek, others: typically charge only cache hit pricing with no write or storage fee.

See Prompt Caching for the full breakdown.

Figures represent median (P50) measurement over the past 72 hours to reflect sustained changes in performance.

Speed

Measured by Output Speed (tokens per second)

Output Speed: Qwen3 235B A22B 2507 Instruct

Output speed: output tokens per second · 10,000 input tokens

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

Figures represent performance of the model's first-party API (e.g. OpenAI for o1) or the median across providers where a first-party API is not available (e.g. Meta's Llama models).

Latency vs. Output Speed

Latency: seconds to first token received · Output speed: output tokens per second · 10,000 input tokens
Most attractive quadrant
Pareto line

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

Time to first token received, in seconds, after API request sent. For reasoning models which share reasoning tokens, this will be the first reasoning token. For models which do not support streaming, this represents time to receive the completion.

Price per token, shown in USD per million tokens. Price is a blend of cache hit, input, and output token prices using the selected ratio (default 7:2:1 cache-input-output).

Figures represent median (P50) measurement over the past 72 hours to reflect sustained changes in performance.

Latency

Measured by Time (seconds) to First Token

Time to First Token

Seconds to first token received · Lower is better · 10,000 input tokens

Time to first token received, in seconds, after API request sent. For reasoning models which share reasoning tokens, this will be the first reasoning token. For models which do not support streaming, this represents time to receive the completion.

Figures represent median (P50) measurement over the past 72 hours to reflect sustained changes in performance.

End-to-End Response Time

Seconds to output 500 tokens, calculated based on time to first token, 'thinking' time for reasoning models, and output speed

End-to-End Response Time

Seconds to output 500 tokens, including reasoning model 'thinking' time · Lower is better · 10,000 input tokens

Seconds to receive a 500 token response. Key components:

  • Input time: Time to receive the first response token
  • Thinking time (only for reasoning models): Time reasoning models spend outputting tokens to reason prior to providing an answer. Amount of tokens based on the average reasoning tokens across a diverse set of 60 prompts (methodology details).
  • Answer time: Time to generate 500 output tokens, based on output speed

For fair comparison, the number of reasoning tokens is standardized across all providers for each model based on the model's representative query token counts.

Figures represent median (P50) measurement over the past 72 hours to reflect sustained changes in performance.

Key Comparison Metrics & API Features

Nebius logoNebius
262k
Open
--
63
1.36
9.29
--
Novita logoNovita
131k
Open
--
40
1.66
14.10
--
DeepInfra logoDeepInfra
262k
Open
--
17
0.60
30.69
--
FriendliAI logoFriendliAI
262k
Open
--
23
2.21
24.17
--
Hyperbolic logoHyperbolic
262k
Open
--
--
--
--
--
Amazon Bedrock logoAmazon Bedrock
256k
Open
--
67
1.37
8.84
--
Parasail logoParasail
262k
Open
--
29
1.47
18.93
--
Scaleway logoScaleway
250k
Open
--
87
1.50
7.26
--
Alibaba Cloud logoAlibaba Cloud
131k
Open
--
56
2.33
11.24
--
Google logoGoogle
262k
Open
--
101
0.79
5.76
--

Frequently Asked Questions

Common questions about Qwen3 235B A22B 2507 Instruct providers

Qwen3 235B A22B 2507 Instruct is available through 10 API providers: Nebius, Novita, DeepInfra (FP8), FriendliAI, Hyperbolic, Amazon, Parasail, Scaleway, Alibaba Cloud, and Google Vertex. Each provider offers different performance characteristics and pricing.

Qwen3 235B A22B 2507 Instruct is currently available through 10 API providers that we benchmark and track.

The fastest providers for Qwen3 235B A22B 2507 Instruct by output speed are Google Vertex (100.7 t/s), Scaleway (86.8 t/s), and Amazon (66.9 t/s). Output speed measures how quickly tokens are generated after the model starts responding.

The providers with the lowest time to first token for Qwen3 235B A22B 2507 Instruct are DeepInfra (FP8) (0.60s), Google Vertex (0.79s), and Nebius (1.36s). Lower latency means faster initial response time.

The most affordable providers for Qwen3 235B A22B 2507 Instruct by blended price are DeepInfra (FP8) ($0.14 per 1M tokens), Novita ($0.14 per 1M tokens), and Parasail ($0.15 per 1M tokens). Blended price uses a 7:2:1 cache hit/input/output token ratio.

The providers with the lowest input token pricing for Qwen3 235B A22B 2507 Instruct are Novita ($0.09 per 1M input tokens), DeepInfra (FP8) ($0.09 per 1M input tokens), and Parasail ($0.15 per 1M input tokens).

The providers with the lowest output token pricing for Qwen3 235B A22B 2507 Instruct are DeepInfra (FP8) ($0.55 per 1M output tokens), Novita ($0.58 per 1M output tokens), and Nebius ($0.60 per 1M output tokens).

Prices for Qwen3 235B A22B 2507 Instruct vary up to 14.7x across providers. The most affordable is DeepInfra (FP8) at $0.14 per 1M tokens, while Hyperbolic charges $2.00 per 1M tokens.

Output speed for Qwen3 235B A22B 2507 Instruct varies significantly across providers. Google Vertex is the fastest at 100.7 t/s, which is 6.1x faster than DeepInfra (FP8) at 16.6 t/s.

9 of 10 providers support JSON mode for Qwen3 235B A22B 2507 Instruct: Nebius, Novita, DeepInfra (FP8), FriendliAI, Hyperbolic, Parasail, Scaleway, Alibaba Cloud, and Google Vertex.

All 10 providers of Qwen3 235B A22B 2507 Instruct support function calling (tool use).

The best provider for Qwen3 235B A22B 2507 Instruct depends on your priorities: Google Vertex offers the highest output speed, DeepInfra (FP8) has the lowest latency, and DeepInfra (FP8) provides the most competitive pricing.

When choosing a provider for Qwen3 235B A22B 2507 Instruct, consider: output speed (for throughput-intensive tasks), latency (for interactive applications requiring quick first responses), pricing (for cost-sensitive workloads), and API features like JSON mode or function calling.

Yes, provider performance can vary over time due to infrastructure changes, load balancing, and updates. We continuously benchmark all providers and display historical performance trends in the "Over Time" charts.

For information about Qwen3 235B A22B 2507 Instruct's intelligence, capabilities, modalities, and how it compares to other models, see the model overview page. View model overview