Digital Mechanics Digital Mechanics Laboratory
ruen

AI model pricing

Prices are in coins per 1 million tokens. An approximate money equivalent appears below.

Claude Fable 5.1 new
The strongest model in the lineup: architecture, deep analysis and work where a mistake costs more than the tokens.		Flagship9,000≈ $2160,000≈ $1421,300≈ $313,000≈ $31
GPT 6 Astra new
OpenAI's flagship: long chains of reasoning, untangling contradictory requirements, mission-critical code.		Flagship6,000≈ $1432,000≈ $76600≈ $17,500≈ $18
Claude Opus 5
Large refactorings, unfamiliar codebases and thorough review — where careful edits matter most.		Flagship5,200≈ $1230,000≈ $71500≈ $18,000≈ $19
Claude Sonnet 5
The everyday all-rounder: complex logic, large edits and tidy code.		Advanced2,500≈ $613,000≈ $31250≈ $13,500≈ $8
Grok 4.6
A fast model for unconventional solutions.		Advanced2,000≈ $57,000≈ $172203,000≈ $7
Kimi K3
Code-specialised: generation, edits and making sense of unfamiliar sections.		Balanced1,700≈ $49,500≈ $231902,500≈ $6
Gemini 3.1 Pro
Multi-step analysis and planning across large volumes of code and documentation.		Advanced1,200≈ $38,000≈ $19250≈ $13,000≈ $7
GPT 5.6 Sol
OpenAI's flagship: long chains of reasoning, untangling contradictory requirements, mission-critical code.		Flagship1,200≈ $38,000≈ $191202,000≈ $5
GPT 5.6 Terra
OpenAI's workhorse.		Advanced1,200≈ $37,500≈ $181201,800≈ $4
GLM 5.2
The best value in the lineup: analysis, code and complex tasks at a fraction of the flagship price.		Balanced800≈ $22,200≈ $51201,000≈ $2
Muse Spark 1.2
Meta's coding-focused model: code generation, refactoring and finding your way around an unfamiliar project — at a fraction of flagship prices.		Balanced700≈ $23,000≈ $7801,200≈ $3
DeepSeek v4 Pro
Economical work on queries, code and algorithms.		Basic500≈ $12,000≈ $5801,500≈ $4
Gemini 3.7 Flash
A high-speed reasoning assistant: fast refactoring, complex query generation, and 1C/BSL coding with massive context support.		Balanced400≈ $12,500≈ $650700≈ $2
MiniMax M3
An inexpensive all-rounder: explanations, code reading and straightforward changes.		Basic400≈ $11,500≈ $480500≈ $1
Gemma 4
The cheapest option: simple transformations, drafts and bulk minor edits.		Basic120600≈ $125170
Claude Fable 5.1 new Flagship

The strongest model in the lineup: architecture, deep analysis and work where a mistake costs more than the tokens.

Input
9,000≈ $21
Output
60,000≈ $142
Cache read
1,300≈ $3
Cache write
13,000≈ $31
GPT 6 Astra new Flagship

OpenAI's flagship: long chains of reasoning, untangling contradictory requirements, mission-critical code.

Input
6,000≈ $14
Output
32,000≈ $76
Cache read
600≈ $1
Cache write
7,500≈ $18
Claude Opus 5 Flagship

Large refactorings, unfamiliar codebases and thorough review — where careful edits matter most.

Input
5,200≈ $12
Output
30,000≈ $71
Cache read
500≈ $1
Cache write
8,000≈ $19
Claude Sonnet 5 Advanced

The everyday all-rounder: complex logic, large edits and tidy code.

Input
2,500≈ $6
Output
13,000≈ $31
Cache read
250≈ $1
Cache write
3,500≈ $8
Grok 4.6 Advanced

A fast model for unconventional solutions.

Input
2,000≈ $5
Output
7,000≈ $17
Cache read
220
Cache write
3,000≈ $7
Kimi K3 Balanced

Code-specialised: generation, edits and making sense of unfamiliar sections.

Input
1,700≈ $4
Output
9,500≈ $23
Cache read
190
Cache write
2,500≈ $6
Gemini 3.1 Pro Advanced

Multi-step analysis and planning across large volumes of code and documentation.

Input
1,200≈ $3
Output
8,000≈ $19
Cache read
250≈ $1
Cache write
3,000≈ $7
GPT 5.6 Sol Flagship

OpenAI's flagship: long chains of reasoning, untangling contradictory requirements, mission-critical code.

Input
1,200≈ $3
Output
8,000≈ $19
Cache read
120
Cache write
2,000≈ $5
GPT 5.6 Terra Advanced

OpenAI's workhorse.

Input
1,200≈ $3
Output
7,500≈ $18
Cache read
120
Cache write
1,800≈ $4
GLM 5.2 Balanced

The best value in the lineup: analysis, code and complex tasks at a fraction of the flagship price.

Input
800≈ $2
Output
2,200≈ $5
Cache read
120
Cache write
1,000≈ $2
Muse Spark 1.2 Balanced

Meta's coding-focused model: code generation, refactoring and finding your way around an unfamiliar project — at a fraction of flagship prices.

Input
700≈ $2
Output
3,000≈ $7
Cache read
80
Cache write
1,200≈ $3
DeepSeek v4 Pro Basic

Economical work on queries, code and algorithms.

Input
500≈ $1
Output
2,000≈ $5
Cache read
80
Cache write
1,500≈ $4
Gemini 3.7 Flash Balanced

A high-speed reasoning assistant: fast refactoring, complex query generation, and 1C/BSL coding with massive context support.

Input
400≈ $1
Output
2,500≈ $6
Cache read
50
Cache write
700≈ $2
MiniMax M3 Basic

An inexpensive all-rounder: explanations, code reading and straightforward changes.

Input
400≈ $1
Output
1,500≈ $4
Cache read
80
Cache write
500≈ $1
Gemma 4 Basic

The cheapest option: simple transformations, drafts and bulk minor edits.

Input
120
Output
600≈ $1
Cache read
25
Cache write
170

Query Console services

coins per use
  • Processor generation (.epf) 1
  • Report generation (.erf) 1

Actual usage depends on context and response size. Cached tokens help reduce the cost of long sessions.