Price Per TokenPrice Per Token
Google
Google
vs
Nousresearch
Nousresearch

Gemma 3 4B vs Hermes 2 Pro - Llama-3 8B

A detailed comparison of pricing, benchmarks, and capabilities

OpenClaw

Deploy OpenClaw in Under 1 Minute We handle hosting, scaling, and maintenance

107 out of our 300 tracked models have had a price change in March.

Get our weekly newsletter on pricing changes, new releases, and tools.

Key Takeaways

Gemma 3 4B wins:

  • Cheaper input tokens
  • Cheaper output tokens
  • Larger context window
  • Faster response time
  • Better at coding
  • Better at math
  • Supports vision

Hermes 2 Pro - Llama-3 8B wins:

  • Higher intelligence benchmark
Price Advantage
Gemma 3 4B
Benchmark Advantage
Gemma 3 4B
Context Window
Gemma 3 4B
Speed
Gemma 3 4B

Pricing Comparison

Price Comparison

MetricGemma 3 4BHermes 2 Pro - Llama-3 8BWinner
Input (per 1M tokens)$0.04$0.14 Gemma 3 4B
Output (per 1M tokens)$0.08$0.14 Gemma 3 4B
Using a 3:1 input/output ratio, Gemma 3 4B is 64% cheaper overall.

Gemma 3 4B Providers

Amazon Bedrock $0.04 (Cheapest)
DeepInfra $0.04 (Cheapest)
Fireworks $0.20

Hermes 2 Pro - Llama-3 8B Providers

Novita $0.14 (Cheapest)

Benchmark Comparison

8
Benchmarks Compared
2
Gemma 3 4B Wins
1
Hermes 2 Pro - Llama-3 8B Wins

Benchmark Scores

BenchmarkGemma 3 4BHermes 2 Pro - Llama-3 8BWinner
Intelligence Index
Overall intelligence score
6.322.1
Coding Index
Code generation & understanding
2.9--
Math Index
Mathematical reasoning
12.7--
MMLU Pro
Academic knowledge
41.722.8
GPQA
Graduate-level science
29.15.7
LiveCodeBench
Competitive programming
11.2--
AIME
Competition math
6.3--
BBH
Big-Bench Hard
-30.7-
Gemma 3 4B shows stronger mathematical reasoning abilities.

Cost vs Quality

X-axis:
Y-axis:
Loading chart...
Gemma 3 4B
Other models

Context & Performance

Context Window

Gemma 3 4B
131,072
tokens
Hermes 2 Pro - Llama-3 8B
8,192
tokens
Max output: 8,192 tokens
Gemma 3 4B has a 94% larger context window.

Speed Performance

MetricGemma 3 4BHermes 2 Pro - Llama-3 8BWinner
Tokens/second35.2 tok/sN/A
Time to First Token1.00sN/A

Capabilities

Feature Comparison

FeatureGemma 3 4BHermes 2 Pro - Llama-3 8B
Vision (Image Input)
Tool/Function Calls
Reasoning Mode
Audio Input
Audio Output
PDF Input
Prompt Caching
Web Search

License & Release

PropertyGemma 3 4BHermes 2 Pro - Llama-3 8B
LicenseOpen SourceOpen Source
AuthorGoogleNousresearch
ReleasedMar 2025May 2024

Gemma 3 4B Modalities

Input
textimage
Output
text

Hermes 2 Pro - Llama-3 8B Modalities

Input
text
Output
text

Related Comparisons

Compare Gemma 3 4B with:

Compare Hermes 2 Pro - Llama-3 8B with:

Frequently Asked Questions

Gemma 3 4B has cheaper input pricing at $0.04/M tokens. Gemma 3 4B has cheaper output pricing at $0.08/M tokens.
Gemma 3 4B scores higher on coding benchmarks with a score of 2.9, compared to Hermes 2 Pro - Llama-3 8B's score of N/A.
Gemma 3 4B has a 131,072 token context window, while Hermes 2 Pro - Llama-3 8B has a 8,192 token context window.
Gemma 3 4B supports vision. Hermes 2 Pro - Llama-3 8B does not support vision.
Advertise with us