HAIL.SO / DISPATCH · 2026-07-27 · COMPARE
FILE: COMPARE · 6 of 6 slots

side by side.

← all costsclear
01

Large language models

4 models
Mistral
Codestral 25.08
×
OpenAI
GPT-5.4 nano
×
Alibaba
Qwen3-Max
×
Perplexity
Sonar
×
Output $/MTok$0.90$1.25$6.00$1.00
Input $/MTok$0.30$0.20$1.20$1.00
Cached input $/MTok$0.0200
Context window131,072400,000262,144128,000
Output cap8,192128,00032,7688,000
Modalitiesin: text / out: textin: text, image / out: textin: text / out: textin: text / out: text
Tool use
Structured output
FamilyCodestralGPT-5.4Qwen 3 MaxSonar
Knowledge cutoff2025-08-312025-06-30
Verified2026-07-192026-07-192026-07-192026-07-19
Sourceprovider page ↗provider page ↗provider page ↗provider page ↗
03

Text to speech

2 models
Rime
Rime Mist v3
×
Cartesia
Sonic 3
×
$/1M chars$30.00$50.00
Voice qualityneuralneural
Voice cloning✓ included
Voice count94
Languagesen42+
SSML support
TTFB37ms
Output formats
Verified2026-07-192026-07-19
Sourceprovider page ↗provider page ↗