HAIL.SO / DISPATCH · 2026-07-02 · COMPARE
FILE: COMPARE · 6 of 6 slots

side by side.

← all costsclear
01

Large language models

3 models
DeepSeek
DeepSeek V4 Flash
×
Perplexity
Sonar
×
OpenAI
OpenAI o4-mini
×
Output $/MTok$0.28$1.00$4.40
Input $/MTok$0.14$1.00$1.10
Cached input $/MTok$0.0028$0.2750
Context window1,048,576128,000200,000
Output cap384,0008,000100,000
Modalitiesin: text / out: textin: text / out: textin: text, image / out: text
Tool use
Structured output
FamilyDeepSeek V4Sonaro-series
Knowledge cutoff2024-06-01
Verified2026-07-022026-07-022026-07-02
Sourceprovider page ↗provider page ↗provider page ↗
02

Speech to text

1 model
Speechmatics
Speechmatics Enhanced
×
$/min$0.002150
$/min batch
Streaming
Realtime
Languages56+
Diarizationincluded
TTFW
Verified2026-07-02
Sourceprovider page ↗
03

Text to speech

2 models
Cartesia
Sonic 2
×
Smallest.ai
Lightning v3.1
×
$/1M chars$50.00$17.50
Voice qualityneuralneural
Voice cloning✓ included✓ included
Voice count217
Languagesen, fr, de, es, pt, zh, ja, koen, hi, es, mr, kn, ta, bn, gu, te, ml, pa, or
SSML support
TTFB90ms200ms
Output formatspcm, mp3, wav, ulaw, alaw
Verified2026-07-022026-07-02
Sourceprovider page ↗provider page ↗