Gemma 3 270M
Aug 2025
Gemma 3 270MCompare inon
Benchmarks of Gemma 3 270M12
BenchmarkByScoreRank
Overall1
Knowledge4
Science2
Coding1
Legacy · saturated or no longer reported4
No benchmark matches
—per task?
self-host onlyprice
—output speed?
32Kcontext
Ranking?
Alternatives?
ModelScoreCostPer taskWhat changes
BetterGPT-6 Lunalow×4.2—$0.005Function calling only with none · Chat Completions+3IOpenK2 Horizon 7Bthinking×4—$0.30 / $1.20Memory (4-bit): 0.1 GB → 4.5 GB
BetterGranite 4.2 3BthinkingOpen+77%—$0.006Memory (4-bit): 0.1 GB → 1.8 GBBetterGemma 4 E4BthinkingOpen+74%—$0.008Memory (4-bit): 0.1 GB → 4 GBBetterGPT-5.6 Lunalow×4.1—$0.01ModelScoreCostPer taskWhat changes
Nothing here yet.
ModelScoreCostPer taskWhat changes
GPT-6 Lunalow×4.2—$0.005Function calling only with none · Chat Completions+3Granite 4.2 3BthinkingOpen+77%—$0.006Memory (4-bit): 0.1 GB → 1.8 GBGemma 4 E4BthinkingOpen+74%—$0.008Memory (4-bit): 0.1 GB → 4 GBGPT-5.6 Lunalow×4.1—$0.01GPT OSS 20BhighOpen+75%—$0.011Memory (4-bit): 0.1 GB → 13 GBModelScoreCostPer taskWhat changes
IK2 Horizon 7BthinkingOpen×4—$0.30 / $1.20Memory (4-bit): 0.1 GB → 4.5 GBIK2 Horizon 3.7BthinkingOpen×3.1—$0.30 / $1.20Memory (4-bit): 0.1 GB → 2.5 GBGemma 4 12BthinkingOpen×2.8—$0.10 / $0.30Memory (4-bit): 0.1 GB → 6 GB
Qwen 3.5 9Bnon-thinkingOpen×2.6—$0.08 / $0.13Memory (4-bit): 0.1 GB → 4.8 GBQwen3.5 4BthinkingOpen×2.6—$0.03 / $0.15Memory (4-bit): 0.1 GB → 2.3 GBModelScoreCostPer taskWhat changes
Gemini 4 Argonhigh×10.3—*$1.99Synchronous output stops at 65,536+2Gemini 3.8 Flashmedium×7.8—*$0.93Sampling parameters are ignored+1Gemini 3.6 Flashhigh×6.6—*$1.60Gemini 3.1 Pro Previewthinking×5.8—$1.30Cannot use both thinking controlsGemini 3.5 Flash Litehigh×4.3—$0.19Gemini 3 Flash Previewnon-thinking×3.5—$0.50 / $3Gemma 4 26B A4BthinkingOpen×3.3—$0.15 / $0.60Memory (4-bit): 0.1 GB → 15 GBGemini 2.5 Prothinking×3.1—$0.33All Google models →Run 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
LicenceGemma termscommercial use with conditions
Benchmarks
all categoriesChat & Knowledge
Humanity's Last Exam3.6#324
AA Omniscience Index-29.3#148
CritPt (physics)0#180
AA Intelligence Index 4.3.2≈5.1#320
AA Omniscience · Accuracy0.9#321
GPQA Diamond22.4#315
AIME 2025 · AA archived2.3#136
Coding
LiveCodeBench · AA archived0.3#153
Agents & Tool Use
τ²-bench · AA9.1#229
Writing & Language
IFBench (instructions)12.1#248
FAQ
How much does Gemma 3 270M cost?›
There is no per-unit API price: self-host only.
Is Gemma 3 270M free to use?›
There is no free API. The weights are open, so you can run it yourself.
How good is Gemma 3 270M, and what is it best at?›
#321 of 334 in Chat & Knowledge, #210 of 239 in Coding, #226 of 242 in Agents & Tool Use, #245 of 245 in Writing & Language.
Is Gemma 3 270M better than Qwen3 1.7B?›
On AA Intelligence Index 4.3.2, Gemma 3 270M scores 5.1 vs 5.2 — about the same.
What is the context window of Gemma 3 270M?›
32K tokens.
When was Gemma 3 270M released? Is there a newer version?›
Released 2025-08-14. It is the current version in its line.
Is Gemma 3 270M open source?›
Yes, the weights are open (gemma licence).
Can I run Gemma 3 270M locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Gemma 3 270M read images?›
No, it accepts text only.
What is Gemma 3 270M?›
Gemma 3 270M is an open-weights language model from Google, released Aug 2025.
What are the best alternatives to Gemma 3 270M?›
GPT-6 Luna (better), K2 Horizon 7B (open weights).