One agent session
1M input tokens and 100K output tokens, no caching.
- Input1M x $0.30 = $0.30
- Output0.1M x $1.50 = $0.15
$0.45per session
Alibaba Cloud (Qwen) - Closed weights
Qwen3-Coder-Flash pricing and specs
Qwen3-Coder-Flash is a closed-weights model from Alibaba Cloud (Qwen) that costs $0.30 per 1M input tokens and $1.50 per 1M output tokens on the first-party API, with a 1M-token context window.
Last verified from alibabacloud.com. All sources
- Input / 1M
- $0.30
- Output / 1M
- $1.50
- Context
- 1M tokens
- Weights
- Closed
Pricing per 1M tokens
| Item | Price / 1M tokens |
|---|---|
| InputPrompts up to 32,000 tokens | $0.30 |
| OutputPrompts up to 32,000 tokens | $1.50 |
| Cache readPrompts up to 32,000 tokens | $0.03 |
| Cache write (5 minutes)Prompts up to 32,000 tokens | $0.375 |
| Long-context inputPrompts over 32,000 tokens | $0.50 |
| Long-context outputPrompts over 32,000 tokens | $2.50 |
Note: International (Singapore) list prices. The text generation page lists it as a legacy model. Four input-length tiers, each billing the whole request: up to 32K $0.3 / $1.5; 32K to 128K $0.5 / $2.5 (recorded as long-context pricing); 128K to 256K $0.8 / $4; 256K to 1M $1.6 / $9.6. cache read/cache write are explicit-cache read and creation for the up-to-32K tier; implicit cache hits are $0.06 in that tier. The alias currently points to qwen3-coder-flash-2025-07-28. The console states context 1M and max output 64K; recorded as 1,000,000 and 65,536.
Specs
- Context window
- 1,000,000 tokens
- Max output
- 65,536 tokens
- Release date
- Provider
- Alibaba Cloud (Qwen)
- API model id
qwen3-coder-flash- Weights
- Closed weights
What Qwen3-Coder-Flash costs in practice
Worked from the standard list rates above, with the math shown so you can swap in your own token counts. These use the base rates, which apply while each request's prompt stays at or under 32,000 tokens; above that, Qwen3-Coder-Flash bills $0.50 input and $2.50 output for the whole request.
A chat app
10,000 requests a day at 2K input and 500 output tokens each.
- Input20M x $0.30 = $6.00
- Output5M x $1.50 = $7.50
$13.50per day
$405.00 per 30-day month
The same session at a 90% cache hit
90% of the input is read from the prompt cache. The one-time cache write premium is not included.
- Cached input0.9M x $0.03 = $0.027
- Uncached input0.1M x $0.30 = $0.03
- Output0.1M x $1.50 = $0.15
$0.207per session
Compare with
| Model | Input / 1M | Output / 1M |
|---|---|---|
| Qwen3-Coder-Flash (this page) | $0.30 | $1.50 |
| Aya Expanse 32BCohere | $0.50 | $1.50 |
| Gemini 3.1 Flash-LiteGoogle | $0.25 | $1.50 |
| GPT-3.5 TurboOpenAI | $0.50 | $1.50 |
| Mistral Large 3Mistral AI | $0.50 | $1.50 |
| Qwen3.6-FlashAlibaba Cloud (Qwen) | $0.25 | $1.50 |
| Qwen3-VL-PlusAlibaba Cloud (Qwen) | $0.20 | $1.60 |
| Qwen3.7-PlusAlibaba Cloud (Qwen) | $0.40 | $1.60 |
| Qwen-PlusAlibaba Cloud (Qwen) | $0.40 | $1.20 |
| Qwen3.5-35B-A3BAlibaba Cloud (Qwen) | $0.25 | $2 |
| Qwen3.5-Omni-FlashAlibaba Cloud (Qwen) | $0.40 | $2.20 |
Related reading
- Run Qwen3.8-Flash-Next on a Gaming PC With StrataRun Qwen3.8-Flash-Next locally with Strata on a 12 GB GPU: hardware needs, speeds by quant, the license catch, and when the hosted API is the better call.
- Terminal-Universe Turns Agent Traces Into Training EnvironmentsTerminal-Universe, Qwen's paper on rebuilding terminal-agent trajectories into reusable environments, and why agent teams should keep replayable workspaces.
- Qwen3.8-27B vs Opus 4.6 Max: Local Agentic Coding BenchmarksQwen3.8-27B is a 27B dense Apache-2.0 model that beats Opus 4.6 Max on SWE-bench Pro (61.7 vs 53.4). Benchmarks, hardware, cost and when to run it.
- Qwen 3.8 Max Ships: 2.4T MoE, 1M Context, $2/$6 per MTok, Open Weights Next WeekQwen 3.8 Max is a 2.4T-parameter MoE with 95B active, a 1M context window and $2/$6 per million tokens on QwenCloud, with open weights promised next week.
- Qwen-UI-Agent Points at the Next GUI Agent RuntimeQwen-UI-Agent, Alibaba's GUI agent report, reads best as a product spec: mobile, desktop, browser, CLI and DeepSearch combined in one stateful agent runtime.
- $500 RL Fine-Tune of a 9B Open Model Beat GPT-5.6 Sol and Claude Opus 4.8 on Catalog ReviewA $500 GRPO fine-tune of Qwen 3.5 9B beat GPT-5.6 Sol and Opus 4.8 on catalog review, hitting 97% accuracy at a fraction of the cost per listing.
FAQ
How much does Qwen3-Coder-Flash cost per million tokens?
Qwen3-Coder-Flash costs $0.30 per 1M input tokens and $1.50 per 1M output tokens on Alibaba Cloud (Qwen)'s first-party API, per alibabacloud.com on October 17, 2026. Per 1M tokens, cache reads cost $0.03. Prompts over 32,000 tokens are billed at $0.50 input and $2.50 output for the whole request.
What is Qwen3-Coder-Flash's context window?
Qwen3-Coder-Flash has a 1,000,000-token context window and returns up to 65,536 output tokens per response.
Are Qwen3-Coder-Flash's weights open?
No. Qwen3-Coder-Flash is a closed-weights model. You call it through Alibaba Cloud (Qwen)'s API as qwen3-coder-flash.
When was Qwen3-Coder-Flash released?
Alibaba Cloud (Qwen) released Qwen3-Coder-Flash on July 29, 2025.
Sources
Last verified from alibabacloud.com.
- https://www.alibabacloud.com/help/en/model-studio/model-pricing
- https://modelstudio.console.alibabacloud.com/ap-southeast-1/model/market/detail/qwen3-coder-flash
- https://www.alibabacloud.com/help/en/model-studio/text-generation-model
- https://www.alibabacloud.com/help/en/model-studio/context-cache
- https://www.alibabacloud.com/help/en/model-studio/newly-released-models
Newsletter
Get model price changes as they ship
One email per week. New model releases, price cuts and what they mean for your bill.