LLM Pricing
ModelsLeaderboardsToolsProvidersReleasesArticles

© 2026 LLM Pricing

About
·Contact
·Privacy
·RSS
← Model list

Qwen3 235B-A22B Thinking 2507

alibaba·alibaba/qwen3-235b-a22b-thinking-2507·GA·Open weights·qwen series
Updated large open Qwen3 MoE thinking model for reasoning, math, coding, and tool use

Specs & pricing

Input / output per 1M tokens
Official price
—
Lowest paid·IO.NETCloud
$0.11 / $0.60
Blended $0.23 · 4.1× spread
1 more $0 channels
Context
128,000
Output limit
16,384
Knowledge cutoff
2025-04
Released / updated
2025-07-25 / 2025-07-25
Capabilities
✓ Reasoning✓ Tool use✓ Structured output✓ TemperatureAttachments
⚠ Providers report capability flags inconsistently
Modalities
Text
License
Apache-2.0
Weights
Hugging Face

Qwen3 235B-A22B Thinking 2507 by alibaba is offered by 12 providers on this page. Public prices are shown for 11 of them. No official API price is listed for it; the prices below come from third-party channels. The lowest paid channel is IO.NET at $0.11 / $0.60 per 1M. Prices vary about 4.1× across the providers that host it. It also has 1 free ($0) channel; free tiers usually carry rate limits, and subscription-covered access bills $0 per token only after the subscription fee.

The context window is 128,000 tokens at the reference host, but hosts report different limits, from 128,000 to 262,144, so the usable window depends on the provider you pick. It supports reasoning, tool use, and structured output. The weights are open, so it can also be self-hosted or served through a gateway of your choice. Providers report the capability flags inconsistently, so verify a specific feature against the host you plan to use. Its training knowledge cuts off at 2025-04.

Available at 12 providers11 with public prices · 1 free

  • ProviderInput / Output
  • ModelScopeCloud
    Qwen/Qwen3-235B-A22B-Thinking-2507
    Free
    Context 262,144 ⚠Output limit 131,072
  • IO.NETCloud
    Qwen/Qwen3-235B-A22B-Thinking-2507
    $0.11 / $0.60
    Cache read $0.055Context 262,144 ⚠Output limit 4,096deprecated
  • submodelGateway
    Qwen/Qwen3-235B-A22B-Thinking-2507
    $0.20 / $0.60
    Context 262,144 ⚠Output limit 131,072
  • NanoGPTGateway
    $0.30 / $0.50
    Cache read $0.15Context 131,072 ⚠Output limit 81,920
  • Eden AIGateway
    deepinfra/Qwen/Qwen3-235B-A22B-Thinking-2507
    $0.23 / $2.30
    Cache read $0.20Context 262,144 ⚠Output limit 81,920
  • QiniuGateway
    $0.29 / $2.90
    Context 262,144 ⚠Output limit 4,096deprecated
  • HeliconeGateway
    qwen3-235b-a22b-thinking
    $0.30 / $2.90
    Context 262,144 ⚠Output limit 81,920
  • Jiekou.AIGateway
    $0.30 / $3.00
    Context 131,072 ⚠Output limit 131,072
  • Hugging FaceCloud
    Qwen/Qwen3-235B-A22B-Thinking-2507
    $0.30 / $3.00
    Context 262,144 ⚠Output limit 81,920
  • Venice AIGateway
    $0.45 / $3.50
    Context 128,000Output limit 16,384
  • OpenRouterGateway
    $0.45 / $3.50
    Context 128,000Output limit 16,384
  • Kilo GatewayGateway
    $0.45 / $3.50
    Context 128,000Output limit 16,384
ProviderTierInputOutputCache readCache writeContextOutput limitStatus
ModelScope
Qwen/Qwen3-235B-A22B-Thinking-2507
CloudFree——262,144 ⚠131,072
IO.NET
Qwen/Qwen3-235B-A22B-Thinking-2507
Cloud$0.11$0.60$0.055$0.22262,144 ⚠4,096deprecated
submodel
Qwen/Qwen3-235B-A22B-Thinking-2507
Gateway$0.20$0.60——262,144 ⚠131,072
NanoGPTGateway$0.30$0.50$0.15—131,072 ⚠81,920
Eden AI
deepinfra/Qwen/Qwen3-235B-A22B-Thinking-2507
Gateway$0.23$2.30$0.20—262,144 ⚠81,920
QiniuGateway$0.29$2.90——262,144 ⚠4,096deprecated
Helicone
qwen3-235b-a22b-thinking
Gateway$0.30$2.90——262,144 ⚠81,920
Jiekou.AIGateway$0.30$3.00——131,072 ⚠131,072
Hugging Face
Qwen/Qwen3-235B-A22B-Thinking-2507
Cloud$0.30$3.00——262,144 ⚠81,920
Venice AIGateway$0.45$3.50——128,00016,384
OpenRouterGateway$0.45$3.50——128,00016,384
Kilo GatewayGateway$0.45$3.50——128,00016,384

Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.

Reasoning control

effort = loweffort = mediumeffort = high

Interleaved thinking (reasoning between tool calls) is declared by 1 of 12 providers.

9 / 12 providers expose no reasoning control (reasoning_options: []).

Your usage cost

Common scenarios
1IO.NET$45.40
2NanoGPT$67.00
3submodel$70.00
4Eden AI$157.40
5Qiniu$202.90
6Helicone$205.00

1 more channels offer $0 (ModelScope); free tiers usually have rate limits and no SLA, excluded from ranking.

The cheapest paid channel is the only channel.

Price historyone sample accumulated per data sync

Input listOutput listMin blended
$0.35$02026-08-052026-09-03
2026-09-03Input list —Output list —Min blended $0.23

Related models

Qwen4 Maxsame series—Qwen 3.6 Plus Uncensoredsame series—Qwen 3.8 27B Uncensoredsame series—Qwen3 235B-A22B Instruct 2507same lab—Qwen3 30B A3Bsame lab—
Data partly from models.dev (MIT)