QED Nano measured on 13 of 104 Chat & Knowledge benchmarks13
BenchmarkByScoreRank
No benchmark matches
Alternatives?
ModelScoreCostPer taskWhat changes
CheaperStep 3.7 FlashunstatedOpenwins 58%-13%$0.20 / $1.15Context: 262K → 256K+1
BetterGPT-5.5xhighwins 61%+2,150%$2.63Prompts with >272K input tokens
OpenQwen3.5 4Bunstatedsame-88%$0.03 / $0.15Memory (4-bit): 2 GB → 2.3 GBCheaperGPT OSS 120BunstatedOpenwins 53%-88%$0.03 / $0.15Context: 262K → 131K+1CheaperQwen 3.5 9BunstatedOpensame-82%$0.08 / $0.13Memory (4-bit): 2 GB → 4.8 GBModelScoreCostPer taskWhat changes
Step 3.7 FlashunstatedOpenwins 58%-13%$0.20 / $1.15Context: 262K → 256K+1GPT OSS 120BunstatedOpenwins 53%-88%$0.03 / $0.15Context: 262K → 131K+1Qwen3.5 4BunstatedOpensame-88%$0.03 / $0.15Memory (4-bit): 2 GB → 2.3 GBQwen 3.5 9BunstatedOpensame-82%$0.08 / $0.13Memory (4-bit): 2 GB → 4.8 GBQwen3.5 2BunstatedOpensame-92%$0.02 / $0.10ModelScoreCostPer taskWhat changes
GPT-5.5xhighwins 61%+2,150%$2.63Prompts with >272K input tokens
Claude Opus 4.8maxwins 60%+1,900%$4.08
Gemini 3.1 Pro Previewunstatedwins 60%+800%$2 / $12Cannot use both thinking controls
Kimi K3unstatedOpenwins 60%+1,100%$3 / $15Memory (4-bit): 2 GB → 1,668 GB+4
DeepSeek V4.1 Flash 1017maxOpenwins 60%+5%$0.30 / $1.20Memory (4-bit): 2 GB → 458 GB+5Run it yourself
estimateFits on✓ 8 GB (phone, laptop) · 8-bit✓ 16 GB · full quality✓ 24 GB GPU · full quality✓ 48 GB GPU · full quality✓ 80 GB GPU · full quality
LicenceApache-2.0free for commercial use
Benchmarks
in Chat & KnowledgeMathArena · AIME 202682.5#24
MathArena · AIME 202577.5#23
MathArena · Apex1.6#23
MathArena · Apex Shortlist22.3#29
MathArena · BRUMO 202587.5#18
MathArena · CMIMC 202561.9#20
MathArena · SMT 202579.7#20
FAQ
How much does QED Nano cost?›
$0.40 per 1M input tokens and $0.80 per 1M output tokens (cheapest host: Featherless).
Is QED Nano free to use?›
There is no free API; the cheapest paid price is $0.40 per 1M input tokens. The weights are open, so you can run it yourself.
How good is QED Nano, and what is it best at?›
#23 of 34 in Chat & Knowledge.
Is QED Nano better than Qwen 3.5 9B?›
On MathArena · Apex Shortlist, QED Nano scores 22.3 vs 29.8 — about the same.
How can I use QED Nano?›
Through the lm-provers API.
What is the context window of QED Nano?›
262K tokens.
When was QED Nano released? Is there a newer version?›
Released 2026-02-12. It is the current version in its line.
Is QED Nano open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run QED Nano locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can QED Nano read images?›
No, it accepts text only.
What is QED Nano?›
QED Nano is an open-weights language model from lm-provers, released Feb 2026.
What are the best alternatives to QED Nano?›
Step 3.7 Flash (cheaper), GPT-5.5 (better), Qwen3.5 4B (open weights).