The Token Tape

daily brief ·

Requesty Cuts Prices 10% Across 40 Models on models.dev

Prices

A blanket 10% price cut swept through requesty's model catalog on models.dev today, touching roughly 40 listings spanning nearly every major lab's models routed through the platform.

Flagship models saw notable drops: gpt-5.5 input fell $5.5→$4.95/MTok with output $33→$29.7 (evt 72514), claude-fable-5@eu input dropped $11→$9.9/MTok with output $55→$49.5 (evt 72511), and gpt-5.4 input went $2.75→$2.475/MTok with output $16.5→$14.85 (evt 72523).

Mid-tier and budget models followed the same -10% pattern: kimi-k2.6 input $0.95→$0.855/MTok (evt 72543), glm-5.1 input $1.4→$1.26/MTok across both standard and @eu variants (evt 72542, evt 72533), grok-4.3 input $1.25→$1.125/MTok (evt 72541), and gemini-3.5-flash input $1.5→$1.35/MTok with output $9→$8.1 (evt 72509).

Smaller/nano tiers also dropped: gpt-5.4-nano input $0.2→$0.18/MTok (evt 72540), nemotron-3-nano-omni input $0.06→$0.054/MTok on both standard and @eu (evt 72525, evt 72524), and mistral-small-2603 input $0.165→$0.1485/MTok on standard and @eu (evt 72507, evt 72508).

Other cuts hit qwen3.6-plus (evt 72526), minimax-m3@eu (evt 72519), kat-coder-pro (evt 72518), step-3.7-flash (evt 72517), grok-4.2-beta (evt 72505), and fugu-ultra, whose input fell $5→$4.5/MTok with output $30→$27 (evt 72506).

events in this brief

On the tape

40 detected events · each links its primary source