# Coding models: quality, price and availability
Source page: [Coding models](https://devforth.io/models-for-code/)
Related agent plan comparison: [Markdown](https://devforth.io/agents-for-code.md) · [Web page](https://devforth.io/agents-for-code/)
Latest model record update: 2026-09-24T10:57:48.133Z

DevForth Coding Score is a relative 0–100 index across the currently published models, based on the listed benchmarks. Average coding task cost uses measured DevForth task tokens and official API prices; it is an estimate, not a subscription charge. DF Coding Price includes the input and cached tokens used to generate 1M output tokens. Lower cost and higher score are better.

## Models

### Claude Opus 5.5 (Anthropic)
- Record updated: 2026-09-23T20:47:07.705Z
- Tier: flagship; limit class: Shared allowance
- Available in: main: claude_code
- DevForth Coding Score (0–100): 99.6
- Benchmark results: Terminal-Bench 4.0 59.6; SciCode 66.9; AutomationBench-AA 69.5; AA-LCR v1.1 84.7; AAII 57.6
- AAII task cost: $5.98
- Average coding task cost: Not listed
- DF Coding Price per 1M output tokens: $66.4 (model token mix)
- Official API price per 1M tokens (USD): input $4, cache read $0.2, output $20, cache write 5m $5, cache write 1h $8
- Speed: unknown latency; unknown output
- Context window: 1M; knowledge cutoff: Jun 2026
- Modality: text, image input → text output

### Claude Fable 5.1 (Anthropic)
- Record updated: 2026-09-23T21:00:33.264Z
- Tier: frontier; limit class: Fable
- Available in: main: claude_code
- DevForth Coding Score (0–100): 90.4
- Benchmark results: Terminal-Bench 4.0 52; SciCode 63.1; AutomationBench-AA 59.4; AA-LCR v1.1 85.3; AAII 53.4
- AAII task cost: $7.63
- Average coding task cost: $2.73
- DF Coding Price per 1M output tokens: $198 (model token mix)
- Official API price per 1M tokens (USD): input $10, cache read $0.25, output $50, cache write 5m $12.5, cache write 1h $20
- Speed: 210.79s latency; 65.1 tokens/s output
- Context window: 1M; knowledge cutoff: Jun 2026
- Modality: text, image input → text output

### GPT-6 Astra (OpenAI)
- Record updated: 2026-09-16T10:41:36.057Z
- Tier: frontier; limit class: Shared allowance
- Available in: main: codex
- DevForth Coding Score (0–100): 88.3
- Benchmark results: Terminal-Bench 4.0 59.1; SciCode 56.5; AutomationBench-AA 68.5; AA-LCR v1.1 80.7; AAII 52.8
- AAII task cost: $3.26
- Average coding task cost: $1.15
- DF Coding Price per 1M output tokens: $379 (model token mix)
- Official API price per 1M tokens (USD): input $10, cache read $1, output $50, cache write 5m $12.5
- Speed: unknown latency; unknown output
- Context window: 1M; knowledge cutoff: Apr 30, 2026
- Modality: text, image input → text output

### GPT-6 Sol (OpenAI)
- Record updated: 2026-09-24T10:57:48.133Z
- Tier: flagship; limit class: Shared allowance
- Available in: main: codex
- DevForth Coding Score (0–100): 81.4
- Benchmark results: Terminal-Bench 4.0 43.9; SciCode 57.6; AutomationBench-AA 61.6; AA-LCR v1.1 83.7; AAII 47.5
- AAII task cost: $1.06
- Average coding task cost: Not listed
- DF Coding Price per 1M output tokens: $81.2 (vendor token mix)
- Official API price per 1M tokens (USD): input $2, cache read $0.2, output $10, cache write 5m $2.5
- Speed: 136.12s latency; 115.9 tokens/s output
- Context window: 872k; knowledge cutoff: Apr 20, 2026
- Modality: text, image input → text output

### Claude Fable 5 (Anthropic)
- Record updated: 2026-09-16T08:02:16.684Z
- Tier: frontier; limit class: Fable
- Available in: main: claude_code, github_copilot
- DevForth Coding Score (0–100): 80.9
- Benchmark results: Terminal-Bench 4.0 42.4; SciCode 61; AutomationBench-AA 54.1; AA-LCR v1.1 82.3; AAII 49.7
- AAII task cost: $8.75
- Average coding task cost: $2.36
- DF Coding Price per 1M output tokens: $232 (model token mix)
- Official API price per 1M tokens (USD): input $10, cache read $1, output $50, cache write 5m $12.5, cache write 1h $20
- Speed: 239.01s latency; 72.6 tokens/s output
- Context window: 1M; knowledge cutoff: Not listed
- Modality: text, image input → text output

### Claude Opus 5 (Anthropic)
- Record updated: 2026-09-16T08:02:10.585Z
- Tier: flagship; limit class: Shared allowance
- Available in: main: claude_code, github_copilot
- DevForth Coding Score (0–100): 79.7
- Benchmark results: Terminal-Bench 4.0 49; SciCode 56.4; AutomationBench-AA 56.6; AA-LCR v1.1 79.3; AAII 50.7
- AAII task cost: $5.86
- Average coding task cost: $0.821
- DF Coding Price per 1M output tokens: $157 (model token mix)
- Official API price per 1M tokens (USD): input $5, cache read $0.5, output $25, cache write 5m $6.25, cache write 1h $10
- Speed: 56.78s latency; 54 tokens/s output
- Context window: 1M; knowledge cutoff: May 2026
- Modality: text, image input → text output

### GPT-5.6 Sol (OpenAI)
- Record updated: 2026-09-16T08:03:02.365Z
- Tier: flagship; limit class: Shared allowance
- Available in: main: codex
- DevForth Coding Score (0–100): 79.3
- Benchmark results: Terminal-Bench 4.0 39.9; SciCode 57.1; AutomationBench-AA 60.1; AA-LCR v1.1 84; AAII 47.1
- AAII task cost: $1.99
- Average coding task cost: $0.350
- DF Coding Price per 1M output tokens: $157 (model token mix)
- Official API price per 1M tokens (USD): input $4, cache read $0.4, output $20, cache write 5m $5
- Speed: unknown latency; unknown output
- Context window: 1.05M; knowledge cutoff: Feb 16, 2026
- Modality: text, image input → text output

### GPT-5.6 Terra (OpenAI)
- Record updated: 2026-09-16T08:03:07.660Z
- Tier: mid; limit class: Shared allowance
- Available in: main: codex
- DevForth Coding Score (0–100): 73.7
- Benchmark results: Terminal-Bench 4.0 35.4; SciCode 55; AutomationBench-AA 59.6; AA-LCR v1.1 83; AAII 42.3
- AAII task cost: $1.4
- Average coding task cost: $0.234
- DF Coding Price per 1M output tokens: $81.4 (model token mix)
- Official API price per 1M tokens (USD): input $2, cache read $0.2, output $12, cache write 5m $2.5
- Speed: unknown latency; unknown output
- Context window: 1.05M; knowledge cutoff: Feb 16, 2026
- Modality: text, image input → text output

### Gemini 3.8 Flash (Google)
- Record updated: 2026-09-16T11:53:14.252Z
- Tier: mid; limit class: Main
- Available in: main: antigravity
- DevForth Coding Score (0–100): 67.9
- Benchmark results: Terminal-Bench 4.0 19.7; SciCode 56.6; AutomationBench-AA 59.9; AA-LCR v1.1 81.3; AAII 41.2
- AAII task cost: $1.24
- Average coding task cost: $0.135
- DF Coding Price per 1M output tokens: $24.3 (vendor token mix)
- Official API price per 1M tokens (USD): input $0.75, cache read $0.075, output $3.75, cache write 5m $0
- Speed: 14.66s latency; 332 tokens/s output
- Context window: 1M; knowledge cutoff: Mar 2026
- Modality: text, image, speech, video input → text output

### GPT-5.4 (OpenAI)
- Record updated: 2026-09-16T08:03:25.711Z
- Tier: mid; limit class: Shared allowance
- Available in: main: codex, github_copilot
- DevForth Coding Score (0–100): 63.3
- Benchmark results: AA-LCR v1.1 82; AAII 39
- AAII task cost: Not listed
- Average coding task cost: $0.254
- DF Coding Price per 1M output tokens: $85.3 (model token mix)
- Official API price per 1M tokens (USD): input $2.5, cache read $0.25, output $15, cache write 5m $0
- Speed: 113.8s latency; 163.4 tokens/s output
- Context window: 1M; knowledge cutoff: Aug 31, 2025
- Modality: text, image input → text output

### GPT-5.5 (OpenAI)
- Record updated: 2026-09-16T08:02:57.058Z
- Tier: flagship; limit class: Shared allowance
- Available in: main: codex, github_copilot
- DevForth Coding Score (0–100): 62.8
- Benchmark results: Terminal-Bench 4.0 14.6; SciCode 55.8; AutomationBench-AA 47.3; AA-LCR v1.1 84.3; AAII 38.6
- AAII task cost: $2.63
- Average coding task cost: $0.490
- DF Coding Price per 1M output tokens: $195 (model token mix)
- Official API price per 1M tokens (USD): input $5, cache read $0.5, output $30, cache write 5m $0
- Speed: 118.46s latency; 71.6 tokens/s output
- Context window: 1.05M; knowledge cutoff: Dec 1, 2025
- Modality: text, image input → text output

### Claude Opus 4.8 (Anthropic)
- Record updated: 2026-09-16T08:02:45.531Z
- Tier: flagship; limit class: Shared allowance
- Available in: main: claude_code, github_copilot
- DevForth Coding Score (0–100): 61.3
- Benchmark results: Terminal-Bench 4.0 21.7; SciCode 54.4; AutomationBench-AA 45.6; AA-LCR v1.1 77.7; AAII 42
- AAII task cost: $4.08
- Average coding task cost: $1.16
- DF Coding Price per 1M output tokens: $128 (model token mix)
- Official API price per 1M tokens (USD): input $5, cache read $0.5, output $25, cache write 5m $6.25, cache write 1h $10
- Speed: 30.05s latency; 62.2 tokens/s output
- Context window: 1M; knowledge cutoff: Not listed
- Modality: text, image input → text output

### Claude Opus 4.7 (Anthropic)
- Record updated: 2026-09-16T08:04:27.973Z
- Tier: flagship; limit class: Shared allowance
- Available in: main: claude_code, github_copilot
- DevForth Coding Score (0–100): 60.7
- Benchmark results: AA-LCR v1.1 78.7; AAII 40.7
- AAII task cost: Not listed
- Average coding task cost: $1.58
- DF Coding Price per 1M output tokens: $149 (vendor token mix)
- Official API price per 1M tokens (USD): input $5, cache read $0.5, output $25, cache write 5m $6.25, cache write 1h $10
- Speed: unknown latency; 49.5 tokens/s output
- Context window: 1M; knowledge cutoff: Jan 2026
- Modality: text, image input → text output

### GPT-5.6 Luna (OpenAI)
- Record updated: 2026-09-16T08:03:14.384Z
- Tier: light; limit class: Shared allowance
- Available in: main: codex
- DevForth Coding Score (0–100): 60.4
- Benchmark results: Terminal-Bench 4.0 11.6; SciCode 53.6; AutomationBench-AA 50.2; AA-LCR v1.1 83.7; AAII 37.5
- AAII task cost: $0.18
- Average coding task cost: $0.0301
- DF Coding Price per 1M output tokens: $6.04 (model token mix)
- Official API price per 1M tokens (USD): input $0.2, cache read $0.02, output $1.2, cache write 5m $0.25
- Speed: unknown latency; unknown output
- Context window: 1.05M; knowledge cutoff: Feb 16, 2026
- Modality: text, image input → text output

### Claude Sonnet 5 (Anthropic)
- Record updated: 2026-09-16T08:02:39.431Z
- Tier: mid; limit class: Shared allowance
- Available in: main: claude_code, github_copilot
- DevForth Coding Score (0–100): 57.1
- Benchmark results: Terminal-Bench 4.0 14.1; SciCode 54.3; AutomationBench-AA 36.5; AA-LCR v1.1 82; AAII 38.4
- AAII task cost: $5.09
- Average coding task cost: $0.969
- DF Coding Price per 1M output tokens: $78.3 (model token mix)
- Official API price per 1M tokens (USD): input $2, cache read $0.2, output $10, cache write 5m $2.5, cache write 1h $4
- Speed: 144.37s latency; 78.5 tokens/s output
- Context window: 1M; knowledge cutoff: Not listed
- Modality: text, image input → text output

### Gemini 3.1 Pro (Preview) (Google)
- Record updated: 2026-09-16T08:03:59.576Z
- Tier: flagship; limit class: Main
- Available in: main: antigravity, github_copilot
- DevForth Coding Score (0–100): 52.5
- Benchmark results: Terminal-Bench 4.0 4; SciCode 58.7; AutomationBench-AA 35.4; AA-LCR v1.1 82; AAII 30.4
- AAII task cost: $0.67
- Average coding task cost: $0.783
- DF Coding Price per 1M output tokens: $66.8 (vendor token mix)
- Official API price per 1M tokens (USD): input $2, cache read $0.2, output $12, cache write 5m $0
- Speed: 25.74s latency; 132.6 tokens/s output
- Context window: 1M; knowledge cutoff: Not listed
- Modality: text, image, speech, video input → text output

### Claude Opus 4.6 (Anthropic)
- Record updated: 2026-09-16T08:04:22.224Z
- Tier: flagship; limit class: Other
- Available in: main: claude_code, github_copilot; other: antigravity
- DevForth Coding Score (0–100): 51.6
- Benchmark results: AA-LCR v1.1 78; AAII 31.9
- AAII task cost: Not listed
- Average coding task cost: $0.581
- DF Coding Price per 1M output tokens: $149 (vendor token mix)
- Official API price per 1M tokens (USD): input $5, cache read $0.5, output $25, cache write 5m $6.25, cache write 1h $10
- Speed: unknown latency; 42.7 tokens/s output
- Context window: 1M; knowledge cutoff: Not listed
- Modality: text, image input → text output

### Gemini 3.5 Flash (Google)
- Record updated: 2026-09-16T08:04:04.799Z
- Tier: mid; limit class: Main
- Available in: main: antigravity, github_copilot
- DevForth Coding Score (0–100): 48.3
- Benchmark results: Terminal-Bench 4.0 6.6; SciCode 53.9; AutomationBench-AA 42.1; AA-LCR v1.1 73.3; AAII 33
- AAII task cost: $1.56
- Average coding task cost: $0.620
- DF Coding Price per 1M output tokens: $49.0 (model token mix)
- Official API price per 1M tokens (USD): input $1.5, cache read $0.15, output $9, cache write 5m $0
- Speed: 20.59s latency; 165.4 tokens/s output
- Context window: 1M; knowledge cutoff: Not listed
- Modality: text, image, speech, video input → text output

### Claude Opus 4.5 (Anthropic)
- Record updated: 2026-09-16T08:04:16.663Z
- Tier: flagship; limit class: Shared allowance
- Available in: main: claude_code, github_copilot
- DevForth Coding Score (0–100): 48.1
- Benchmark results: AA-LCR v1.1 77.3; AAII 29.1
- AAII task cost: Not listed
- Average coding task cost: Not listed
- DF Coding Price per 1M output tokens: $149 (vendor token mix)
- Official API price per 1M tokens (USD): input $5, cache read $0.5, output $25, cache write 5m $6.25, cache write 1h $10
- Speed: 15.04s latency; 53.1 tokens/s output
- Context window: 200k; knowledge cutoff: Aug 2025
- Modality: text, image input → text output

### Gemini 3 Flash (Preview) (Google)
- Record updated: 2026-09-16T08:03:53.848Z
- Tier: light; limit class: Main
- Available in: main: antigravity, github_copilot
- DevForth Coding Score (0–100): 46.3
- Benchmark results: AA-LCR v1.1 78; AAII 26.3
- AAII task cost: Not listed
- Average coding task cost: Not listed
- DF Coding Price per 1M output tokens: $17.5 (model token mix)
- Official API price per 1M tokens (USD): input $0.5, cache read $0.05, output $3, cache write 5m $0
- Speed: 7.6s latency; 166.4 tokens/s output
- Context window: 1M; knowledge cutoff: Jan 2025
- Modality: text, image, speech, video input → text output

### Claude Sonnet 4.6 (Anthropic)
- Record updated: 2026-09-16T08:02:34.113Z
- Tier: mid; limit class: Other
- Available in: main: claude_code, github_copilot; other: antigravity
- DevForth Coding Score (0–100): 41.4
- Benchmark results: Terminal-Bench 4.0 3; SciCode 50.1; AutomationBench-AA 20.1; AA-LCR v1.1 80; AAII 30.5
- AAII task cost: $2.49
- Average coding task cost: Not listed
- DF Coding Price per 1M output tokens: $98.1 (model token mix)
- Official API price per 1M tokens (USD): input $3, cache read $0.3, output $15, cache write 5m $3.75, cache write 1h $6
- Speed: 1.16s latency; 42.7 tokens/s output
- Context window: 1M; knowledge cutoff: Not listed
- Modality: text, image input → text output

### GPT-5.4 mini (OpenAI)
- Record updated: 2026-09-16T08:03:20.288Z
- Tier: light; limit class: Shared allowance
- Available in: main: codex, github_copilot
- DevForth Coding Score (0–100): 39.6
- Benchmark results: Terminal-Bench 4.0 2; SciCode 52.1; AutomationBench-AA 26.2; AA-LCR v1.1 77; AAII 24.6
- AAII task cost: $0.41
- Average coding task cost: $0.183
- DF Coding Price per 1M output tokens: $16.7 (model token mix)
- Official API price per 1M tokens (USD): input $0.75, cache read $0.075, output $4.5, cache write 5m $0
- Speed: 11.87s latency; 183.2 tokens/s output
- Context window: 400k; knowledge cutoff: Aug 31, 2025
- Modality: text, image input → text output

### Kimi-K2.7-Code (Moonshot AI)
- Record updated: 2026-09-16T08:04:33.569Z
- Tier: mid; limit class: Shared allowance
- Available in: main: github_copilot
- DevForth Coding Score (0–100): 38.3
- Benchmark results: Terminal-Bench 4.0 1; SciCode 47.8; AutomationBench-AA 24.5; AA-LCR v1.1 79.3; AAII 26.3
- AAII task cost: $0.54
- Average coding task cost: Not listed
- DF Coding Price per 1M output tokens: $40.9 (total token mix)
- Official API price per 1M tokens (USD): input $0.95, cache read $0.19, output $4, cache write 5m $0
- Speed: 2.92s latency; 53.9 tokens/s output
- Context window: 256k; knowledge cutoff: Not listed
- Modality: text, image, video input → text output

### GPT-5.4 nano (OpenAI)
- Record updated: 2026-09-16T08:04:11.113Z
- Tier: light; limit class: Shared allowance
- Available in: main: github_copilot
- DevForth Coding Score (0–100): 28.6
- Benchmark results: Terminal-Bench 4.0 0.5; SciCode 47.2; AutomationBench-AA 6; AA-LCR v1.1 76.7; AAII 21.2
- AAII task cost: $0.18
- Average coding task cost: Not listed
- DF Coding Price per 1M output tokens: $8.37 (vendor token mix)
- Official API price per 1M tokens (USD): input $0.2, cache read $0.02, output $1.25, cache write 5m $0
- Speed: 4.74s latency; 159 tokens/s output
- Context window: 400k; knowledge cutoff: Aug 31, 2025
- Modality: text, image input → text output

### Claude Sonnet 4.5 (Anthropic)
- Record updated: 2026-09-16T08:03:43.453Z
- Tier: mid; limit class: Shared allowance
- Available in: main: claude_code, github_copilot
- DevForth Coding Score (0–100): 27.2
- Benchmark results: Terminal-Bench 4.0 0; SciCode 45.7; AutomationBench-AA 14.1; AA-LCR v1.1 72.3; AAII 21.2
- AAII task cost: $0.56
- Average coding task cost: Not listed
- DF Coding Price per 1M output tokens: $89.5 (vendor token mix)
- Official API price per 1M tokens (USD): input $3, cache read $0.3, output $15, cache write 5m $3.75, cache write 1h $6
- Speed: 1.36s latency; 41.2 tokens/s output
- Context window: 1M; knowledge cutoff: Jul 2025
- Modality: text, image input → text output

### Claude Haiku 4.5 (Anthropic)
- Record updated: 2026-09-16T08:02:51.162Z
- Tier: light; limit class: Shared allowance
- Available in: main: claude_code, github_copilot
- DevForth Coding Score (0–100): 21.6
- Benchmark results: Terminal-Bench 4.0 0; SciCode 42.2; AutomationBench-AA 3.2; AA-LCR v1.1 74.3; AAII 17.6
- AAII task cost: $0.21
- Average coding task cost: $0.159
- DF Coding Price per 1M output tokens: $29.4 (model token mix)
- Official API price per 1M tokens (USD): input $1, cache read $0.1, output $5, cache write 5m $1.25, cache write 1h $2
- Speed: 0.81s latency; 93.4 tokens/s output
- Context window: 200k; knowledge cutoff: Jul 2025
- Modality: text, image input → text output

### Gemini 2.5 Pro (Google)
- Record updated: 2026-09-16T08:03:48.751Z
- Tier: flagship; limit class: Shared allowance
- Available in: main: github_copilot
- DevForth Coding Score (0–100): 20.2
- Benchmark results: Terminal-Bench 4.0 0; SciCode 46.3; AutomationBench-AA 2.2; AA-LCR v1.1 69; AAII 16.7
- AAII task cost: $0.23
- Average coding task cost: Not listed
- DF Coding Price per 1M output tokens: $44.3 (vendor token mix)
- Official API price per 1M tokens (USD): input $1.25, cache read $0.125, output $10, cache write 5m $0
- Speed: 22.54s latency; 144.8 tokens/s output
- Context window: 1M; knowledge cutoff: Jan 1, 2025
- Modality: text, image, speech, video input → text output

### GPT-5 mini (OpenAI)
- Record updated: 2026-09-16T08:03:31.749Z
- Tier: light; limit class: Shared allowance
- Available in: main: github_copilot
- DevForth Coding Score (0–100): 19.3
- Benchmark results: Terminal-Bench 4.0 0; SciCode 39; AutomationBench-AA 6.5; AA-LCR v1.1 72.3; AAII 17.4
- AAII task cost: $0.05
- Average coding task cost: Not listed
- DF Coding Price per 1M output tokens: $10.9 (vendor token mix)
- Official API price per 1M tokens (USD): input $0.25, cache read $0.025, output $2, cache write 5m $0
- Speed: 85.3s latency; 103.1 tokens/s output
- Context window: 400k; knowledge cutoff: May 31, 2024
- Modality: text, image input → text output

### GPT-OSS 120B (OpenAI)
- Record updated: 2026-09-16T08:04:39.641Z
- Tier: light; limit class: Other
- Available in: other: antigravity
- DevForth Coding Score (0–100): 0.0
- Benchmark results: Terminal-Bench 4.0 0; SciCode 34; AutomationBench-AA 0.2; AA-LCR v1.1 52; AAII 12.3
- AAII task cost: $0.11
- Average coding task cost: Not listed
- DF Coding Price per 1M output tokens: $39.1 (vendor token mix)
- Official API price per 1M tokens (USD): input $0.15, cache read $0.15, output $0.6
- Speed: 0.94s latency; 266 tokens/s output
- Context window: 131k; knowledge cutoff: May 2024
- Modality: text input → text output

### Raptor mini (GitHub)
- Record updated: 2026-09-15T18:37:13.586Z
- Tier: light; limit class: Shared allowance
- Available in: main: github_copilot
- DevForth Coding Score (0–100): —
- Benchmark results: Not listed
- AAII task cost: Not listed
- Average coding task cost: Not listed
- DF Coding Price per 1M output tokens: $7.23 (total token mix)
- Official API price per 1M tokens (USD): input $0.25, cache read $0.025, output $2, cache write 5m $0
- Speed: unknown latency; unknown output
- Context window: 400k; knowledge cutoff: May 30, 2024
- Modality: text, image input → text output

### MAI-Code-1-Flash (Microsoft)
- Record updated: 2026-09-15T18:37:13.282Z
- Tier: light; limit class: Shared allowance
- Available in: main: github_copilot
- DevForth Coding Score (0–100): —
- Benchmark results: Not listed
- AAII task cost: Not listed
- Average coding task cost: Not listed
- DF Coding Price per 1M output tokens: $20.2 (total token mix)
- Official API price per 1M tokens (USD): input $0.75, cache read $0.075, output $4.5, cache write 5m $0
- Speed: unknown latency; unknown output
- Context window: Not listed; knowledge cutoff: Not listed
- Modality: text input → text output

### Claude Opus 4.8 Fast (Anthropic)
- Record updated: 2026-07-16T00:00:00.000Z
- Tier: flagship; limit class: Shared allowance
- Available in: main: claude_code, github_copilot
- DevForth Coding Score (0–100): —
- Benchmark results: Not listed
- AAII task cost: Not listed
- Average coding task cost: Not listed
- DF Coding Price per 1M output tokens: $299 (vendor token mix)
- Official API price per 1M tokens (USD): input $10, cache read $1, output $50, cache write 5m $12.5, cache write 1h $20
- Speed: unknown latency; unknown output
- Context window: 1M; knowledge cutoff: Not listed
- Modality: text, image input → text output

This file is regenerated from current data on every request. "Record updated" and "Limits last confirmed" reflect stored update and confirmation dates, not the time this file was fetched. If your agent supports monitoring, you may ask it to notify you when a limit changes.
Agent plan data: https://devforth.io/agents-for-code.md