LLM Comparison

DeepSeek V3.1 Terminus vs DeepSeek V4 Pro

Side-by-side specs, pricing & capabilities · Updated August 2026

Price vs Intelligence

5547393123$0.35$0.47$0.60$0.72$0.85Avg Price per M tokensIntelligenceDeepSeek V3.1 Terminus33.2 / $0.50DeepSeek V4 Pro45.4 / $0.65

Add to comparison

2/6 models
Same tier:
 DeepSeekDeepSeek V3.1 TerminusDeepSeekDeepSeek V4 Pro
OrganizationDeepSeekDeepSeekDeepSeekDeepSeek
OpenTools Score
33
66.4
45
69.6
FamilyDeepSeekDeepSeek V4
StatusCurrentCurrent
Release DateSep 2025Aug 2026
Context Window164K tokens1.0M tokens
Input Price$0.21/M tokens$0.43/M tokens
Output Price$0.79/M tokens$0.87/M tokens
Pricing NotesCache read: $0.1300/M tokensDeepSeek API pricing currently lists $0.003625 cache-hit input, $0.435 cache-miss input, and $0.87 output per 1M tokens for deepseek-v4-pro. DeepSeek announced peak/off-peak pricing effective August 16, 2026, with off-peak prices at half the peak rate.
Capabilities
textcode
reasoningcodingtool-usestructured-outputlong-contextagenticresponses-api
Max Output—393K tokens
API Identifierdeepseek/deepseek-v3.1-terminusdeepseek-v4-pro
Benchmarks
MMLU-Pro
83.6deepseek
—
GPQA Diamond
75.1deepseek
—
AIME 2025
53.7deepseek
—
LiveCodeBench
52.9deepseek
—
SWE-bench Verified
66deepseek
—
Terminal-Bench Hard
31.8deepseek
—
HLE
8.4deepseek
—
Artificial Analysis Intelligence Index v4.1—
44artificial-analysis
Artificial Analysis Intelligence Index v4.1.1—
53artificial-analysis
 View DeepSeek V3.1 TerminusView DeepSeek V4 Pro

Cost Calculator

Enter your expected monthly token usage to compare costs.

ModelInputOutputTotal / movs Best
DeepSeek V3.1 TerminusCheapest$0.21$0.40$0.61—
DeepSeek V4 Pro$0.44$0.44$0.87+44%

DeepSeek

DeepSeek V3.1 Terminus

DeepSeek V3.1 Terminus is a large language model from DeepSeek. Supports up to 163,840 token context window. Achieves 87.1% on MMLU. Available from $0.21/M input tokens.

DeepSeek

DeepSeek V4 Pro

DeepSeek V4 Pro 0813 is an open-weight Mixture-of-Experts reasoning model for long-context coding, tool use, and agentic workloads, with native Responses API support through the DeepSeek API.

More Comparisons

Flagship

Claude Opus 5 vs Gemini 3.6 Flash

Flagship

Claude Opus 5 vs GPT-5.6 Sol

Flagship

Claude Opus 5 vs Palmyra X5

Flagship

Claude Opus 5 vs Jamba Large 1.7

Flagship

Claude Opus 5 vs Command A

Flagship

Claude Opus 5 vs Sonar Deep Research

Looking for more AI models?

Browse All LLMs