What are the token costs and operational benchmarks for Composer 2.5?

Composer 2.5 is priced at $1.80 per million input tokens and $7.20 per million output tokens. It features a 262.144k token context window, an average response latency of 180ms TTFT, and achieves 74.6% on SWE-bench Verified and 83.5% on MMLU-Pro.
Verified daily via automated API latency tests and official documentation.
Input / 1M $1.80
Output / 1M $7.20
Context Limit 262.144k
TTFT Latency 180ms
SWE-bench 74.6%
Throughput 130 t/s

Architectural Overview

Frontier foundation model developed by Anysphere / Cursor featuring Full-Repository Diff & Refactor Specialist architecture.

Optimal Production Use Cases

Precision multi-file git diff synthesis, automated code review, IDE background indexing, and full-stack repo migration.

Interactive Monthly Token Economics & ROI Forecaster

Model your expected production workload across prompt (input) and completion (output) tokens.

Live Calculation Engine
10.0M Tokens
100K 50M 250M 500M+
2.5M Tokens
100K 10M 50M 100M+

Estimated Monthly Spend

Composer 2.5 $5.45

$1.8/1M in · $7.2/1M out

GPT-5.6 Luna $67.50

$0.18/1M in · $0.72/1M out

Projected Monthly Cost Reduction
$62.05 / mo
(91.9% lower cost)

Head-to-Head Comparisons Involving Composer 2.5

Versus Comparison

Composer 2.5 vs Claude 3.5 Haiku

Claude 3.5 Haiku is 55.6% cheaper for input tokens ($0.80 vs. $1.80 per 1M tokens) and $4.00 vs. $7.20 for output tokens (2.2x cost difference). In terms of operational performance, Claude 3.5 Haiku delivers faster response latency with 140ms TTFT (40ms faster than Composer 2.5). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 40.6% for Claude 3.5 Haiku.

Versus Comparison

Composer 2.5 vs Claude 3.5 Sonnet

Composer 2.5 is 40.0% cheaper for input tokens ($1.80 vs. $3.00 per 1M tokens) and $7.20 vs. $15.00 for output tokens (1.7x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (140ms faster than Claude 3.5 Sonnet). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 63.7% for Claude 3.5 Sonnet.

Versus Comparison

Composer 2.5 vs Claude 3.7 Sonnet

Composer 2.5 is 40.0% cheaper for input tokens ($1.80 vs. $3.00 per 1M tokens) and $7.20 vs. $15.00 for output tokens (1.7x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (470ms faster than Claude 3.7 Sonnet). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 70.3% for Claude 3.7 Sonnet.

Versus Comparison

Composer 2.5 vs Claude Opus 5

Composer 2.5 is 64.0% cheaper for input tokens ($1.80 vs. $5.00 per 1M tokens) and $7.20 vs. $25.00 for output tokens (2.8x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (160ms faster than Claude Opus 5). Claude Opus 5 leads coding benchmarks at 82.4% SWE-bench vs. 74.6% for Composer 2.5.

Versus Comparison

Composer 2.5 vs Codestral 25.01

Codestral 25.01 is 83.3% cheaper for input tokens ($0.30 vs. $1.80 per 1M tokens) and $0.90 vs. $7.20 for output tokens (6.0x cost difference). In terms of operational performance, Codestral 25.01 delivers faster response latency with 150ms TTFT (30ms faster than Composer 2.5). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 44.2% for Codestral 25.01.

Versus Comparison

Composer 2.5 vs DeepSeek-R1

DeepSeek-R1 is 69.4% cheaper for input tokens ($0.55 vs. $1.80 per 1M tokens) and $2.19 vs. $7.20 for output tokens (3.3x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (1620ms faster than DeepSeek-R1). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 49.2% for DeepSeek-R1.

Versus Comparison

Composer 2.5 vs DeepSeek-V3

DeepSeek-V3 is 92.2% cheaper for input tokens ($0.14 vs. $1.80 per 1M tokens) and $0.28 vs. $7.20 for output tokens (12.9x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (160ms faster than DeepSeek-V3). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 42.0% for DeepSeek-V3.

Versus Comparison

Composer 2.5 vs DeepSeek-V4 Flash

DeepSeek-V4 Flash is 92.2% cheaper for input tokens ($0.14 vs. $1.80 per 1M tokens) and $0.56 vs. $7.20 for output tokens (12.9x cost difference). In terms of operational performance, DeepSeek-V4 Flash delivers faster response latency with 150ms TTFT (30ms faster than Composer 2.5). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 62.4% for DeepSeek-V4 Flash.

Versus Comparison

Composer 2.5 vs Fable 5

Composer 2.5 is 10.0% cheaper for input tokens ($1.80 vs. $2.00 per 1M tokens) and $7.20 vs. $8.00 for output tokens (1.1x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (60ms faster than Fable 5). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 58.0% for Fable 5.

Versus Comparison

Composer 2.5 vs Gemini 2.0 Flash

Gemini 2.0 Flash is 94.4% cheaper for input tokens ($0.10 vs. $1.80 per 1M tokens) and $0.40 vs. $7.20 for output tokens (18.0x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (200ms faster than Gemini 2.0 Flash). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 48.0% for Gemini 2.0 Flash.

Versus Comparison

Composer 2.5 vs Gemini 3.7 Flash

Gemini 3.7 Flash is 95.6% cheaper for input tokens ($0.08 vs. $1.80 per 1M tokens) and $0.32 vs. $7.20 for output tokens (22.5x cost difference). In terms of operational performance, Gemini 3.7 Flash delivers faster response latency with 75ms TTFT (105ms faster than Composer 2.5). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 68.2% for Gemini 3.7 Flash.

Versus Comparison

Composer 2.5 vs GLM 5.3 Flash

GLM 5.3 Flash is 91.7% cheaper for input tokens ($0.15 vs. $1.80 per 1M tokens) and $0.50 vs. $7.20 for output tokens (12.0x cost difference). In terms of operational performance, GLM 5.3 Flash delivers faster response latency with 130ms TTFT (50ms faster than Composer 2.5). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 54.2% for GLM 5.3 Flash.

Versus Comparison

Composer 2.5 vs OpenAI GPT-4o

Composer 2.5 is 28.0% cheaper for input tokens ($1.80 vs. $2.50 per 1M tokens) and $7.20 vs. $10.00 for output tokens (1.4x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (100ms faster than OpenAI GPT-4o). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 38.8% for OpenAI GPT-4o.

Versus Comparison

Composer 2.5 vs GPT-5.6 Luna

GPT-5.6 Luna is 90.0% cheaper for input tokens ($0.18 vs. $1.80 per 1M tokens) and $0.72 vs. $7.20 for output tokens (10.0x cost difference). In terms of operational performance, GPT-5.6 Luna delivers faster response latency with 90ms TTFT (90ms faster than Composer 2.5). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 48.5% for GPT-5.6 Luna.

Versus Comparison

Composer 2.5 vs GPT-5.6 Sol

Composer 2.5 is 77.5% cheaper for input tokens ($1.80 vs. $8.00 per 1M tokens) and $7.20 vs. $32.00 for output tokens (4.4x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (240ms faster than GPT-5.6 Sol). GPT-5.6 Sol leads coding benchmarks at 79.5% SWE-bench vs. 74.6% for Composer 2.5.

Versus Comparison

Composer 2.5 vs GPT-5.6 Terra

GPT-5.6 Terra is 16.7% cheaper for input tokens ($1.50 vs. $1.80 per 1M tokens) and $6.00 vs. $7.20 for output tokens (1.2x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (30ms faster than GPT-5.6 Terra). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 65.4% for GPT-5.6 Terra.

Versus Comparison

Composer 2.5 vs Grok 3

Composer 2.5 is 40.0% cheaper for input tokens ($1.80 vs. $3.00 per 1M tokens) and $7.20 vs. $15.00 for output tokens (1.7x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (670ms faster than Grok 3). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 58.5% for Grok 3.

Versus Comparison

Composer 2.5 vs xAI Grok 4.6

Composer 2.5 is 10.0% cheaper for input tokens ($1.80 vs. $2.00 per 1M tokens) and $7.20 vs. $6.00 for output tokens (1.1x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (100ms faster than xAI Grok 4.6). xAI Grok 4.6 leads coding benchmarks at 76.8% SWE-bench vs. 74.6% for Composer 2.5.

Versus Comparison

Composer 2.5 vs Llama 3.3 70B Instruct

Llama 3.3 70B Instruct is 90.0% cheaper for input tokens ($0.18 vs. $1.80 per 1M tokens) and $0.40 vs. $7.20 for output tokens (10.0x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (240ms faster than Llama 3.3 70B Instruct). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 38.8% for Llama 3.3 70B Instruct.

Versus Comparison

Composer 2.5 vs Mistral Large 2

Composer 2.5 is 10.0% cheaper for input tokens ($1.80 vs. $2.00 per 1M tokens) and $7.20 vs. $6.00 for output tokens (1.1x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (370ms faster than Mistral Large 2). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 39.0% for Mistral Large 2.

Versus Comparison

Composer 2.5 vs OpenAI o1

Composer 2.5 is 88.0% cheaper for input tokens ($1.80 vs. $15.00 per 1M tokens) and $7.20 vs. $60.00 for output tokens (8.3x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (670ms faster than OpenAI o1). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 48.9% for OpenAI o1.

Versus Comparison

Composer 2.5 vs o3-mini

o3-mini is 38.9% cheaper for input tokens ($1.10 vs. $1.80 per 1M tokens) and $4.40 vs. $7.20 for output tokens (1.6x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (1020ms faster than o3-mini). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 49.3% for o3-mini.

Versus Comparison

Composer 2.5 vs Microsoft Phi-4 (14B)

Microsoft Phi-4 (14B) is 93.3% cheaper for input tokens ($0.12 vs. $1.80 per 1M tokens) and $0.36 vs. $7.20 for output tokens (15.0x cost difference). In terms of operational performance, Microsoft Phi-4 (14B) delivers faster response latency with 110ms TTFT (70ms faster than Composer 2.5). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 42.1% for Microsoft Phi-4 (14B).

Versus Comparison

Composer 2.5 vs Qwen 2.5 72B Instruct

Qwen 2.5 72B Instruct is 80.6% cheaper for input tokens ($0.35 vs. $1.80 per 1M tokens) and $0.40 vs. $7.20 for output tokens (5.1x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (240ms faster than Qwen 2.5 72B Instruct). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 44.0% for Qwen 2.5 72B Instruct.

Versus Comparison

Composer 2.5 vs Qwen 2.5 Max

Qwen 2.5 Max is 84.4% cheaper for input tokens ($0.28 vs. $1.80 per 1M tokens) and $0.84 vs. $7.20 for output tokens (6.4x cost difference). In terms of operational performance, Composer 2.5 delivers faster response latency with 180ms TTFT (300ms faster than Qwen 2.5 Max). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 44.2% for Qwen 2.5 Max.

Versus Comparison

Composer 2.5 vs Qwen 3.8 Flash Next

Qwen 3.8 Flash Next is 93.3% cheaper for input tokens ($0.12 vs. $1.80 per 1M tokens) and $0.48 vs. $7.20 for output tokens (15.0x cost difference). In terms of operational performance, Qwen 3.8 Flash Next delivers faster response latency with 120ms TTFT (60ms faster than Composer 2.5). Composer 2.5 leads coding benchmarks at 74.6% SWE-bench vs. 56.8% for Qwen 3.8 Flash Next.

Frequently Asked Questions & Query Fan-Out

How much does Composer 2.5 cost per 1M tokens?

Composer 2.5 costs $1.80 per million prompt (input) tokens and $7.20 per million completion (output) tokens.

What is the context window for Composer 2.5?

Composer 2.5 supports a maximum context window of 262,144 tokens, with a maximum single-generation output of 32,768 tokens.

What are the primary use cases for Composer 2.5?

Precision multi-file git diff synthesis, automated code review, IDE background indexing, and full-stack repo migration.