Anthropic Claude
Claude Haiku 5.5
$0.20 / 1M tokens
cheapest at Anthropic · $0.10 in · $0.50 out
Wins 3 of 4
Head to head · prices updated Nov 2, 2026
Claude Haiku 5.5 vs Gemma 3 4B IT
Claude Haiku 5.5 costs 4.0× as much as Gemma 3 4B IT at their cheapest providers (blended 3 input : 1 output tokens).
Google Gemini and Gemma
Gemma 3 4B IT
$0.05 / 1M tokens
cheapest at Amazon Bedrock · $0.04 in · $0.08 out
Wins 1 of 4
- Lower priceGemma 3 4B IT
- Higher accuracy (GPQA)Claude Haiku 5.5
- Longer contextClaude Haiku 5.5
- More providersClaude Haiku 5.5
Verdict
Gemma 3 4B IT is the better deal on price
Claude Haiku 5.5 costs 4.0× as much as Gemma 3 4B IT at their cheapest providers (blended 3 input : 1 output tokens).
Claude Haiku 5.5 scores higher on GPQA Diamond: 89.6% (max reasoning) vs 23.2%.
So the choice is a trade-off: Claude Haiku 5.5 for the hardest reasoning, Gemma 3 4B IT when cost matters more than the last few points of accuracy.
Claude Haiku 5.5 accepts longer prompts (1000K vs 131K tokens).
Your monthly bill
What would each cost you?
Gemma 3 4B IT saves you $1.44 a month ($17.28 a year) for this workload.
Side by side
| Claude Haiku 5.5 | Gemma 3 4B IT | |
|---|---|---|
| Cheapest blended price / 1M | $0.20 | $0.05 |
| Input / output per 1M | $0.10 / $0.50 | $0.04 / $0.08 |
| Accuracy (GPQA Diamond) | 89.6% (max reasoning) | 23.2% |
| Max context | 1M | 131K |
| Providers we track | 4 | 2 |
Standard list prices at each model's cheapest provider we track; blended = 3 input : 1 output tokens. GPQA Diamond from Epoch AI (CC BY 4.0), best measured reasoning setting, named next to each score. Methodology.
Where to buy each one
Claude Haiku 5.5
- Anthropic$0.20
- Amazon Bedrock$0.20
- Google Cloud Vertex AI$0.20
- Microsoft Azure AI Foundry$0.20
Gemma 3 4B IT
- Amazon Bedrock$0.05
- DeepInfra$0.0625
Questions people ask
Is Claude Haiku 5.5 cheaper than Gemma 3 4B IT?
Claude Haiku 5.5 costs 4.0× as much as Gemma 3 4B IT at their cheapest providers (blended 3 input : 1 output tokens).
Which is more accurate, Claude Haiku 5.5 or Gemma 3 4B IT?
On GPQA Diamond (graduate-level science questions, measured by Epoch AI) Claude Haiku 5.5 scores 89.6% (max reasoning) and Gemma 3 4B IT 23.2%. One benchmark is not the whole picture; test both on your own prompts.
How much does Claude Haiku 5.5 cost per 1M tokens?
From $0.10 input and $0.50 output per 1M tokens at Anthropic, the cheapest of 4 providers we track.
How much does Gemma 3 4B IT cost per 1M tokens?
From $0.04 input and $0.08 output per 1M tokens at Amazon Bedrock, the cheapest of 2 providers we track.
Swap a model