← Back to all models
C

CodeGemma 7B

by Google DeepMind
Efficient Free & Open Source 🏆 Ranked #78 of 85
55.4
Overall Score
out of 100
About

Google DeepMind's code-specialised 7B model built on Gemma, pre-trained on 500B+ tokens of code. Excels at code completion, generation, and natural language to code conversion.

Key Metrics
Context Window
8K
tokens
Avg Response
380
milliseconds
Input Cost
$0.04
per million tokens
Output Cost
$0.04
per million tokens
Arena ELO
1145
Chatbot Arena rating
MT-Bench
7.7
out of 10
Benchmark Scores
MMLU
54.0%
HumanEval
82.0%
MATH
50.0%
GPQA
25.0%
MT-Bench
77.0/10
Capability Profile
Strengths & Limitations
Strengths
✓ Code generation ✓ Open source ✓ Fast ✓ Low memory ✓ Multi-language coding
Limitations
⚠ Code-focused ⚠ Below general LLMs on text tasks ⚠ Older model
Ideal Use Cases
Code completion IDE integration Code generation Bug fixing Code explanation
Model Details
Provider Google DeepMind
Released 2024-04-09
Type Free & Open Source
Multimodal No
Tier Efficient
Global rank #78 / 85
Pricing (USD)
Input tokens $0.04/M
Output tokens $0.04/M
Per 1,000 tokens ≈ $0.0000 input / $0.0000 output
All Benchmarks
MMLU 54.0%
HumanEval 82.0%
MATH 50.0%
GPQA 25.0%
MT-Bench 7.7/10
Arena ELO 1145
Compare this model View Rankings

You might also consider