Qwen3 Coder 480B A35B
Qwen3 Coder 480B A35BRankingin non-thinking mode?
Alternatives?
CheaperMiMo v2.6 ProthinkingOpen×10.9-12%$0.13Memory (4-bit): 288 GB → 615 GB+4
BetterGPT-6 Lunalow×4.5-67%$0.005Function calling only with none · Chat Completions+3
CheaperMiniMax M3.1 Flashthinking×10.6-57%$0.22Complete model response must be appended
CheaperClaude Haiku 5.5max×9.8-67%$0.21Remove sampling parameters+5
CheaperDeepSeek V4.1 Flash 1017maxOpen×9.5-15%$0.26Memory (4-bit): 288 GB → 458 GB+5MiMo v2.6 ProthinkingOpen×10.9-12%$0.13Memory (4-bit): 288 GB → 615 GB+4MiniMax M3.1 Flashthinking×10.6-57%$0.22Complete model response must be appendedClaude Haiku 5.5max×9.8-67%$0.21Remove sampling parameters+5DeepSeek V4.1 Flash 1017maxOpen×9.5-15%$0.26Memory (4-bit): 288 GB → 458 GB+5
GLM 5.3 FlashmaxOpen×8.8-61%$0.28Defaults to max · Self-hosted+4GPT-6 Lunalow×4.5-67%$0.005Function calling only with none · Chat Completions+3
Granite 4.2 3BthinkingOpen+46%-91%$0.006Context: 262K → 128KGPT-5.6 Lunalow×4.4-27%$0.01GPT OSS 20BhighOpen×2.4-94%$0.011Context: 262K → 131K
Mistral Small 4thinkingOpen×2.2-57%$0.015MiMo v2.6 ProthinkingOpen×10.9-12%$0.13Memory (4-bit): 288 GB → 615 GB+4
Step 5 PreviewthinkingOpen×10.2+132%$1.03Memory (4-bit): 288 GB → 363 GB+3
Muse Spark OpenxhighOpen×9.7+225%$0.84Memory (4-bit): 288 GB → 493 GB+2DeepSeek V4.1 Flash 1017maxOpen×9.5-15%$0.26Memory (4-bit): 288 GB → 458 GB+5GLM 5.3maxOpen×9.3+250%$2.01Memory (4-bit): 288 GB → 452 GB+4Qwen3.8 Max (0902)thinking×9.5—$5.41Qwen3.8 2.4T A95BthinkingOpen×9.3—$2.16Memory (4-bit): 288 GB → 1,468 GB+5Qwen3.7 Maxthinking×8.9—$1.15Qwen3.8 FlashthinkingOpen×8.4—$0.37max_tokens range · QwenCloud+5Qwen3.7 Plusthinking×7.8—$0.22Qwen3.6 Max Previewthinking×6.8—$1.30 / $7.80Include the word JSON · QwenCloudQwen3.5 397B A17BthinkingOpen×6.4—$0.47Qwen3 Max Thinkingthinking×6.2—$1.20 / $6Context: 262K → 256KAll Alibaba (Qwen) models →Run it yourself
estimateBenchmarks
all categoriesFAQ
How much does Qwen3 Coder 480B A35B cost?›
$0.22 per 1M input tokens and $1.80 per 1M output tokens (cheapest host: Google).
Is Qwen3 Coder 480B A35B free to use?›
There is no free API; the cheapest paid price is $0.22 per 1M input tokens. The weights are open, so you can run it yourself.
How good is Qwen3 Coder 480B A35B, and what is it best at?›
#290 of 334 in Chat & Knowledge, #11 of 19 in Coding, #121 of 242 in Agents & Tool Use, #96 of 195 in Writing & Language.
Is Qwen3 Coder 480B A35B better than Qwen3 VL 4B (Reasoning)?›
On Humanity's Last Exam, Qwen3 Coder 480B A35B scores 4.5 vs 4.6 — about the same.
How can I use Qwen3 Coder 480B A35B?›
Through the API of AWS Bedrock, DeepInfra, Featherless, Google.
What is the context window of Qwen3 Coder 480B A35B?›
262K tokens, with up to 66K output tokens.
When was Qwen3 Coder 480B A35B released? Is there a newer version?›
Released 2025-07-22. It is the current version in its line.
Is Qwen3 Coder 480B A35B open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run Qwen3 Coder 480B A35B locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Qwen3 Coder 480B A35B read images?›
No, it accepts text only.
What is Qwen3 Coder 480B A35B?›
Qwen3 Coder 480B A35B is an open-weights language model from Alibaba (Qwen), released Jul 2025.
What are the best alternatives to Qwen3 Coder 480B A35B?›
MiMo v2.6 Pro (cheaper), GPT-6 Luna (better).