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

side by side.

← all costsclear
01

Large language models

3 models
OpenAI
GPT-4o
×
Alibaba
Qwen3-Max
×
DeepSeek
DeepSeek V3
×
Output $/MTok$10.00$6.00$0.28
Input $/MTok$2.50$1.20$0.14
Cached input $/MTok$1.2500$0.0028
Context window128,000262,1441,048,576
Output cap16,38432,768384,000
Modalitiesin: text, image / out: textin: text / out: textin: text / out: text
Tool use
Structured output
FamilyGPT-4oQwen 3 MaxDeepSeek V3
Knowledge cutoff2023-10-012025-06-30
Verified2026-07-192026-07-192026-07-19
Sourceprovider page ↗provider page ↗provider page ↗
02

Speech to text

3 models
Rev.ai
Rev.ai Reverb
×
Deepgram
Flux (Multilingual)
×
Speechmatics
Speechmatics Enhanced
×
$/min$0.003300$0.007800$0.002150
$/min batch
Streaming
Realtime
Languagesenen, es, fr, de, hi, ru, pt, ja, it, nl56+
Diarizationincludedincluded
TTFW
Verified2026-07-192026-07-192026-07-19
Sourceprovider page ↗provider page ↗provider page ↗