Models

DeepSeek V4 Pro (Reasoning, Max Effort)

DeepSeek V4 Pro is a high-compute reasoning model optimized for complex logical, mathematical, and coding tasks. It represents the 'Max Effort' variant, utilizing significant computational resources to maximize accuracy and depth in its reasoning chains.

Reasoning
Input / 1M tokens
$1.74
Output / 1M tokens
$3.48
Output tokens/s
31.3
First-token seconds
1.2s
Supported plans
37

Benchmark history

Evaluations

9

TAU2

Measured May 14, 2026Source

Score

0.96

Terminalbench Hard

Measured May 14, 2026Source

Score

0.46

Lcr

Measured May 14, 2026Source

Score

0.66

Ifbench

Measured May 14, 2026Source

Score

0.76

Scicode

Measured May 14, 2026Source

Score

0.5

Hle

Measured May 14, 2026Source

Score

0.36

Gpqa

Measured May 14, 2026Source

Score

0.89

Artificial Analysis Coding Index

Measured May 14, 2026Source

Score

47.5

Artificial Analysis Intelligence Index

Measured May 14, 2026Source

Score

51.5

CODPL speed

Provider ranking

1

Plan availability

Products and plans that support this model

10
Apertis Coding Plan

Apertis Coding Plan

Apertis Coding Plan is a subscription-based AI coding service providing unified access to 30+ AI models (GPT-5.4, Claude Opus 4.6, Gemini 3.1 Pro, and more) through a single API key. Designed for developers using coding agents like Claude Code, Cursor, Cline, and OpenCode, it offers predictable monthly pricing, free prompt caching, auto-failover, and quota-based billing across OpenAI, Anthropic, Google, and other providers.

OpenCode Go

OpenCode Go

OpenCode Go is a low-cost monthly subscription that provides reliable access to a curated set of powerful open-source coding models, such as GLM-5.1, Kimi K2.6, and DeepSeek V4 Pro, for use with AI coding agents like OpenCode. Priced at $10/month with a $5 first month, it offers generous request limits to support developers.

Discussion

Thinking... Make sure you are connected to GitHub server