HAIL.SO / DISPATCH · 2026-07-23 · COMPARE
FILE: COMPARE · 6 of 6 slots

side by side.

← all costsclear
01

Large language models

4 models
Mistral
Codestral 25.08
×
Alibaba
Qwen3.7-Plus
×
OpenAI
GPT-5.4
×
xAI
Grok 4.3
×
Output $/MTok$0.90$1.60$15.00$2.50
Input $/MTok$0.30$0.40$2.50$1.25
Cached input $/MTok$0.2500$0.2000
Context window131,0721,048,5761,050,0001,000,000
Output cap8,19265,536128,0001,000,000
Modalitiesin: text / out: textin: text, image / out: textin: text, image / out: textin: text, image / out: text
Tool use
Structured output
FamilyCodestralQwen 3.7 PlusGPT-5.4Grok 4
Knowledge cutoff2025-08-31
Verified2026-07-192026-07-192026-07-192026-07-19
Sourceprovider page ↗provider page ↗provider page ↗provider page ↗
02

Speech to text

1 model
Cartesia
Ink Whisper
×
$/min$0.003000
$/min batch$0.001500
Streaming
Realtime
Languages99+
Diarizationunsupported
TTFW
Verified2026-07-19
Sourceprovider page ↗
03

Text to speech

1 model
Rime
Rime Mist v3
×
$/1M chars$30.00
Voice qualityneural
Voice cloning
Voice count94
Languagesen
SSML support
TTFB37ms
Output formats
Verified2026-07-19
Sourceprovider page ↗