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
Ministral 3 14B
×
Alibaba
Qwen3.7-Plus
×
OpenAI
OpenAI o3
×
OpenAI
GPT-5
×
Output $/MTok$0.20$1.60$8.00$10.00
Input $/MTok$0.20$0.40$2.00$1.25
Cached input $/MTok$0.5000$0.1250
Context window262,1441,048,576200,000400,000
Output cap8,19265,536100,000128,000
Modalitiesin: text, image / out: textin: text, image / out: textin: text, image / out: textin: text, image / out: text
Tool use
Structured output
FamilyMinistralQwen 3.7 Pluso-seriesGPT-5
Knowledge cutoff2024-06-012024-09-30
Verified2026-07-192026-07-192026-07-192026-07-19
Sourceprovider page ↗provider page ↗provider page ↗provider page ↗
02

Speech to text

1 model
AssemblyAI
Whisper-Streaming
×
$/min$0.005000
$/min batch
Streaming
Realtime
Languages99+
Diarizationextra-cost
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 ↗