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

side by side.

← all costsclear
01

Large language models

2 models
DeepSeek
DeepSeek V3
×
DeepSeek
DeepSeek R1
×
Output $/MTok$0.28$0.28
Input $/MTok$0.14$0.14
Cached input $/MTok$0.0028$0.0028
Context window1,048,5761,048,576
Output cap384,000384,000
Modalitiesin: text / out: textin: text / out: text
Tool use
Structured output
FamilyDeepSeek V3DeepSeek R1
Knowledge cutoff
Verified2026-07-272026-07-27
Sourceprovider page ↗provider page ↗
02

Speech to text

2 models
Cartesia
Ink Whisper
×
Speechmatics
Speechmatics Enhanced
×
$/min$0.003000$0.002150
$/min batch$0.001500
Streaming
Realtime
Languages99+56+
Diarizationunsupportedincluded
TTFW
Verified2026-07-272026-07-19
Sourceprovider page ↗provider page ↗
03

Text to speech

2 models
Cartesia
Sonic 3.5
×
Deepgram
Deepgram Aura 2
×
$/1M chars$50.00$30.00
Voice qualityneuralneural
Voice cloning✓ included
Voice count
Languages42+en, es, de, fr, nl, it, ja
SSML support
TTFB90ms
Output formatsraw/pcm_f32le, raw/pcm_s16le, raw/pcm_mulaw, raw/pcm_alaw, wav, mp3wav, mp3, linear16, mulaw, alaw, opus
Verified2026-07-272026-07-19
Sourceprovider page ↗provider page ↗