โ† Back to Model Hub/SIDE-BY-SIDE REVIEW
SHARE THIS:

Claude 3 Sonnet vs DeepSeek V3.1 Terminus

How do these models stack up? Below is an expert side-by-side comparison of specifications, context window capacity, live pricing per million tokens, and standardized benchmark scores for Claude 3 Sonnet and DeepSeek V3.1 Terminus.

Anthropic

Claude 3 Sonnet

Claude 3 Sonnet was Anthropic's balanced standard model in the first Claude 3 generation, optimized for a combination of response speed and general-purpose capabilities across writing, summarization, and question answering. Released in March 2024 alongside Opus and Haiku, it occupied the mid-tier positioning in the Claude 3 lineup. Claude 3 Sonnet has since been deprecated, and current use cases requiring similar capability-to-cost balance are best served by the Claude 4 Sonnet series, which offers substantially improved reasoning, context handling, and instruction following at competitive pricing.

View Full Specs
DeepSeek

DeepSeek V3.1 Terminus

DeepSeek-V3.1 Terminus is an update to [DeepSeek V3.1](/deepseek/deepseek-chat-v3.1) that maintains the model's original capabilities while addressing issues reported by users, including language consistency and agent capabilities, further optimizing the model's...

View Full Specs

Technical Specifications

SpecificationClaude 3 SonnetDeepSeek V3.1 Terminus
ProviderAnthropicDeepSeek
Context Window200,000 tokens163,840 tokens
Agent SuitabilityN/AN/A
Time to First Token (TTFT)N/AN/A
Deployment ModelN/Aself hostable
Production Stabilitystablestable
API AvailableYesYes
Released Date2024-03-042025-09-22

API Pricing Comparison

Input Price per Million Tokens

Claude 3 Sonnet

$3.00

DeepSeek V3.1 Terminus

$0.27

Output Price per Million Tokens

Claude 3 Sonnet

$15.00

DeepSeek V3.1 Terminus

$0.95

Want to test both models live?

Run side-by-side prompt prompts in our dynamic Sandbox. Check execution speeds, latency metrics, and compute actual costs in real-time.

Benchmark Performance Metrics

Scores show the raw performance percentages verified across key evaluation suites. Higher bars indicate superior accuracy and capability in that domain.

MMLUGeneral knowledge & multi-task understanding
7900.0%vsN/A
Claude 3 Sonnet
DeepSeek V3.1 Terminus
HumanEvalPython coding & logic synthesis
7300.0%vsN/A
Claude 3 Sonnet
DeepSeek V3.1 Terminus
MATHComplex mathematical problem solving
4100.0%vsN/A
Claude 3 Sonnet
DeepSeek V3.1 Terminus
GPQAGraduate-level expert reasoning
3500.0%vsN/A
Claude 3 Sonnet
DeepSeek V3.1 Terminus
HellaSwagCommonsense reasoning and inference
8900.0%vsN/A
Claude 3 Sonnet
DeepSeek V3.1 Terminus
MT-BenchMulti-turn conversation flow quality
820.0%vsN/A
Claude 3 Sonnet
DeepSeek V3.1 Terminus

Claude 3 Sonnet Quirks & Gotchas

No developer gotchas reported.

DeepSeek V3.1 Terminus Quirks & Gotchas

No developer gotchas reported.