Qwen3 Next 80B A3B Thinking
alibaba/qwen3-next-80b-a3b-thinkingA chat model with a 262K-token context window, at $0.1575 per million input tokens and $1.26 per million output tokens.
Details
- Maker
- alibaba
- Kind
- Chat
- Context window
- 262,144 tokens
- Max output
- 262,144 tokens
- Input
- $0.1575 per million tokens
- Output
- $1.26 per million tokens
- Capabilities
- Reasoning, tool use
- Takes
- Text
- Returns
- Text
- Zero data retention
- Yes, on every provider
- No training on requests
- Yes, on every provider
- Released
- 2025-09-11
- Newer model
- alibaba/qwen3.8-omni-flashReleased later by the same maker, costs no more, and lists the same capabilities. Compare the two
Description
Qwen3-Next uses a highly sparse MoE design: 80B total parameters, but only ~3B activated per inference step. Experiments show that, with global load balancing, increasing total expert parameters while keeping activated experts fixed steadily reduces training loss.Compared to Qwen3’s MoE (128 total experts, 8 routed), Qwen3-Next expands to 512 total experts, combining 10 routed experts + 1 shared expert — maximizing resource usage without hurting performance. The Qwen3-Next-80B-A3B-Thinking excels at complex reasoning tasks — outperforming higher-cost models like Qwen3-30B-A3B-Thinking-2507 and Qwen3-32B-Thinking, outpeforming the closed-source Gemini-2.5-Flash-Thinking on multiple benchmarks, and approaching the performance of our top-tier model Qwen3-235B-A22B-Thinking-2507.
Call it
curl https://phantom.codes/v1/chat/completions \
-H "Authorization: Bearer $PHANTOM_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "alibaba/qwen3-next-80b-a3b-thinking",
"messages": [
{"role": "user", "content": "Hello"}
]
}'Set PHANTOM_KEY to your key. Any OpenAI client works with the base URL https://phantom.codes/v1.