Benchmarks of Nanbeige4 3B Thinking 25114
BenchmarkByScoreRank
Knowledge1
Writing2
Legacy · saturated or no longer reported1
No benchmark matches
—per task?
self-host onlyprice
—output speed?
66Kcontext
Ranking?
Alternatives?
ModelScoreCostPer taskWhat changes
BetterClaude Opus 5unstated×2.7—$5 / $25
OpenGemma 4 12Bunstated+55%—$0.10 / $0.30Memory (4-bit): 2 GB → 6 GBBetterClaude Fable 5.1unstated×2.7—$10 / $50Replace forced tool_choice+5BetterClaude Opus 5.5unstated×2.7—$4 / $20No forced tool use+5BetterClaude Fable 5unstated×2.6—$10 / $50ModelScoreCostPer taskWhat changes
Nothing here yet.
ModelScoreCostPer taskWhat changes
Claude Opus 5unstated×2.7—$5 / $25Claude Fable 5.1unstated×2.7—$10 / $50Replace forced tool_choice+5Claude Opus 5.5unstated×2.7—$4 / $20No forced tool use+5Claude Fable 5unstated×2.6—$10 / $50
Grok 4.7unstated×2.6—$2 / $6Reasoning cannot be disabled+4ModelScoreCostPer taskWhat changes
Gemma 4 12BunstatedOpen+55%—$0.10 / $0.30Memory (4-bit): 2 GB → 6 GBGemma 3 12BunstatedOpen+33%—$0.05 / $0.10Memory (4-bit): 2 GB → 6.1 GBGemma 3 4BunstatedOpen+9%—$0.04 / $0.05Memory (4-bit): 2 GB → 2.2 GB
Mistral NemounstatedOpen-9%—$0.15 / $0.15Memory (4-bit): 2 GB → 6.1 GB
Qwen3 4BunstatedOpen-46%—$0.40 / $0.80Context: 66K → 33KRun it yourself
estimateFits on✓ 8 GB (phone, laptop) · full quality✓ 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
all categoriesChat & Knowledge
EQ-Bench 3728.1#49
Agents & Tool Use
BFCL v451.4#12
Writing & Language
EQ-Bench Creative Writing v3837.4#85
FAQ
How much does Nanbeige4 3B Thinking 2511 cost?›
There is no per-unit API price: self-host only.
Is Nanbeige4 3B Thinking 2511 free to use?›
There is no free API. The weights are open, so you can run it yourself.
How good is Nanbeige4 3B Thinking 2511, and what is it best at?›
#49 of 53 in Chat & Knowledge, #12 of 39 in Agents & Tool Use, #89 of 96 in Writing & Language.
Is Nanbeige4 3B Thinking 2511 better than Qwen3 Next 80B A3B Instruct?›
On EQ-Bench Longform Creative Writing 1.11, Nanbeige4 3B Thinking 2511 scores 31.5 vs 31.5 — about the same.
What is the context window of Nanbeige4 3B Thinking 2511?›
66K tokens.
When was Nanbeige4 3B Thinking 2511 released? Is there a newer version?›
Released 2025-11-20. It is the current version in its line.
Is Nanbeige4 3B Thinking 2511 open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run Nanbeige4 3B Thinking 2511 locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Nanbeige4 3B Thinking 2511 read images?›
No, it accepts text only.
What is Nanbeige4 3B Thinking 2511?›
Nanbeige4 3B Thinking 2511 is an open-weights language model from Nanbeige, released Nov 2025.
What are the best alternatives to Nanbeige4 3B Thinking 2511?›
Claude Opus 5 (better), Gemma 4 12B (open weights).