AiCostCompare

Compare models

Pick up to 4 models for a side-by-side spec sheet. Hover any dotted metric name for what the test measures. Best value in each row is highlighted.

Claude Sonnet 5DeepSeek V4 Flash
Claude Sonnet 5

Anthropic

DeepSeek V4 Flash

DeepSeek

Pricing
Input $/1M$2.00$0.094
Output $/1M$10.00$0.188
Blended $/1M (3:1)$4.00$0.117
Cache read $/1M$0.20$0.019
Cache write $/1M$2.50
Quality indices
Intelligence Index53.440.3
Coding Index71.556.2
Math Index
Agentic Index46.731.1
Speed & latency
Output speed (tok/s)88115
Time to first token93.55s0.86s
Time to first answer token93.55s49.64s
Benchmarks (Artificial Analysis)
GPQA Diamond91.1%89.4%
Humanity's Last Exam39.6%32.1%
SciCode53.6%44.9%
IFBench79.2%
AA-LCR (Long Context Reasoning)70.7%63.0%
Terminal-Bench Hard35.6%
Terminal-Bench 2.180.5%61.8%
τ²-Bench (Telecom)95.0%
τ³-Bench Banking28.2%22.9%
Design Arena (Elo)
Website1,3141,235
Web apps1,305
Full-stack1,292
Android native1,255
UI components1,3171,205
Data viz1,2811,159
SVG1,2391,202
3D1,3211,248
Game dev1,3451,254
Agentic game dev1,241
Godot game dev1,257
HTML slides1,233
Python→PPTX slides1,252
Code categories1,3121,239
ASCII art1,2301,157
Specs
Context window1M1.05M
Max output tokens128K
Input modalitiestext, image, filetext
TokenizerClaudeDeepSeek