Deepgram

Flux TTS 'free until 9/14, then $0.0450/1k chars' promo expired as scheduled; a new promo appeared: 'Buy Flux TTS, get matching credits' (spend $1 get $1 in credits, up to $500) on Pay As You Go and Growth, through Dec 31, 2026.

Flux TTS free-access promo period extended: was free through Sept 12, 2026 (standard pricing from Sept 13); now free through Sept 14, 2026 (standard pricing from Sept 15).

Added new Flux TTS model to Text-to-Speech, free through Sept 12, 2026 (up to 45 concurrent streams); then $0.045/1k chars PAYG, $0.0405/1k chars Growth. Also free in Voice Agent API through Sept 12, then billed at BYO TTS rates.

Added FAQ entries disclosing a 10% overage fee for usage beyond the Growth plan's included capacity, plus guidance on how Growth plan overages are billed

Speech-to-Text model lineup restructured: "Flux" split into separate "Flux English" and new "Flux Multilingual" line items (Multilingual priced at $0.0078/min Pay-As-You-Go, $0.0068/min Growth); Nova-1 & 2, Enhanced, and Base models removed from the main public pricing table (a new FAQ entry "Are older models (Nova-2, Enhanced, Base) still available?" was added, but their pricing is no longer shown in the comparison table); new "Limited-time promotional rates on streaming" discount reduced streaming prices for Flux English, Nova-3 Monolingual, and Nova-3 Multilingual (old prices now shown struck through next to lower promotional prices, e.g. Flux English Pay-As-You-Go $0.0077/min → $0.0065/min) Speech-to-Text REST API concurrency limit cut in half for both the Pay As You Go plan and the Growth plan: "Up to 100 for the REST API" reduced to "Up to 50 for the REST API"

STT model lineup overhauled: Flux renamed to Flux English; Flux Multilingual added; Nova-1 & 2, Enhanced, Base removed; limited-time promotional streaming rates noted.

Speech-to-Text REST API concurrency limit reduced from 100 to 50 for both Pay As You Go and Growth plans.

Voice Agent API price reductions across multiple tiers [Standard PAYG: $0.0800->$0.075/min (-6.25%); Standard Growth: $0.0700->$0.068/min (-2.86%); Custom BYO LLM PAYG: $0.0700->$0.056/min (-20%); Custom BYO LLM Growth: $0.0600->$0.059/min (-1.67%); Advanced Growth: $0.1500->$0.146/min (-2.67%)]

Voice Agent API prices changed across all tiers: Standard PAYGO $0.0800->$0.075, Custom-BYO LLM PAYGO $0.0700->$0.056, Advanced PAYGO $0.1600->$0.163. Enterprise column removed from pricing table. Concurrency limits increased for STT WSS (50->150/225), TTS REST+WSS (15->45/60), and Voice Agent API WSS (15->45/60) for Pay As You Go and Growth plans respectively.

Growth: Speech-to-Text WSS API 50→225 (4.5x) Pay As You Go: Speech-to-Text WSS API 50→150 (3x)

Flux model pricing introduced after promotional period ended: $0.0077/min for Pay As You Go and $0.0065/min for Growth plan (was 'Free through October')

Nova-3 (English) model renamed to Nova-3 (Monolingual) Keyterm Prompting add-on price decreased from $0.0020/min to $0.0013/min (Pay As You Go, 35% decrease) and from $0.0017/min to $0.0012/min (Growth, 29% decrease)

Speech-to-text concurrency limits clarified: Now explicitly separates 100 concurrent requests for pre-recorded models and 50 concurrent requests for streaming (previously grouped as '100 concurrent requests for Deepgram speech-to-text models') Nova-1 and Nova-2 models consolidated into single 'Nova-1 & 2' tier for Pre-Recorded Speech to Text

Enterprise plan minimum price disclosed: $10k+/year (previously labeled 'Exclusive' with no price shown) Growth plan pricing changed from '$4k-10k/year' range to '$4k+/year' open-ended minimum (removed upper bound)

Entity Detection feature added to Speech-to-Text Pre-Recorded and Streaming at $0.0013/min (Pay As You Go) and $0.0011/min (Growth) Growth plan Aura TTS batch API increased from 2 to 3 concurrent requests, 480 to 720 requests/min, and 120 to 180 concurrent conversations (implicit baseline)

Enterprise plan support description changed from 'Premium SLAs, dedicated support teams, and email support' to 'Paid Support plans available' Added self-service form link for requesting higher rate limits for Aura text-to-speech API (available in Pay As You Go and Growth plans)

Aura text-to-speech concurrency limits added: up to 2 concurrent requests and 480 requests/min for Pay As You Go plan Audio Intelligence Summarization pricing changed from per-minute ($0.0044/min Pay As You Go) to token-based ($0.0003/1k input tokens, $0.0006/1k output tokens)