PromptsZone Logo

PromptsZone

AI news & analysis

New Models

GPT-6.2 Standard pricing per million tokens: Astra $8 input and $40 output; Sol $2 input and $10 output, for requests up to 272,000 input tokens. Price comparison, not performance.

GPT-6.2 Astra and Sol reach the API: pricing and a migration trap to check

November 17th

OpenAI adds GPT-6.2 Astra and Sol to the API. Compare token prices and context limits, and check the conversation changes that can discard reasoning state.

Read more
New Models
Tencent headquarters in an unaltered corporate photograph supplied for editorial use.

Tencent Hy4 flash arrives on TokenHub: prices, limits and integration checks

November 16th

Hy4 flash is listed on Tencent TokenHub with lower token prices than Hy4. Check API limits, temporary discounts and a practical migration test.

Read more
New Models
Launch scorecard comparing Gemini 4 Hydrogen, Claude Sonnet 5.5 and GPT-6.1 Terra on AutomationBench-AA, HLE-Diamond and Vending-Bench 2, using Google's compilation of independent leaderboard results.

Gemini 4 Hydrogen launches: pricing, access and six benchmark comparisons

November 12th

Google launches Gemini 4 Hydrogen. Compare six benchmarks with Claude Sonnet 5.5 and GPT-6.1 Terra, plus API prices, rollout and the million-token limit.

Read more
New Models
Three developer-reported English Clean benchmark panels compare ZONOS3-Mini embedding with Qwen3-TTS 0.6B and 1.7B on word error rate, speaker similarity and predicted audio quality; 1,279 prompts per model.

ZONOS3-Mini brings voice cloning to browsers: a matched speech benchmark comparison

November 12th

Zyphra releases ZONOS3-Mini for local and browser speech generation. Compare matched English voice-cloning metrics, model access and licensing.

Read more
New Models
Developer-reported word error rates across 102 FLEURS languages: Scribe v3 12.4%, Scribe v2 17.0%, Gemini 3.5 Transcribe 18.5%; lower is better.

Scribe v3 and v3 Live launch: speech benchmarks and the batch-versus-streaming choice

November 12th

ElevenLabs releases Scribe v3 and v3 Live. Compare matched-language transcription scores, real-time speaker labels and practical migration checks.

Read more
New Models
AlphaGenome 2 TenK10K comparison: 30% versus 16% recall at 90% precision; 43% versus 23% recall at 90% sign accuracy, compared with AlphaGenome 1-sc.

AlphaGenome 2 launches with cell-specific DNA predictions: benchmarks and research access

November 11th

Google DeepMind’s AlphaGenome 2 adds cell-specific variant predictions. Compare its GTEx and TenK10K benchmarks, API access and research limits.

Read more
New Models
Two Google-reported accuracy charts compare Gemini Decision 1, TypeSafe Jev 1.14, GPT-6 Luna and Gemini 3.5 Flash-Lite on JevBench and When2Call.

Gemini Decision 1 launches for fast multimodal choices, with a four-model scorecard

November 11th

Google's Gemini Decision 1 evaluates fixed choices across text and media. Compare its launch benchmarks, pricing, access and practical API limits.

Read more
New Models
DeepL-supplied 2025 product screenshot showing a multilingual video meeting and translated text; product context, not a demonstration of Aure's audio output.

DeepL launches Aure to carry tone and pace into voice translation

November 11th

DeepL's Aure voice model launches in eight languages. Understand its automatic rollout, expression-matching approach and what to check in a multilingual pilot.

Read more
New Models
Original diagram showing Robotics ER 2 planning, Gemini Robotics 2.5 motor control and robot hardware. Includes the official Gemini family logo.

Gemini Robotics 2.5 reaches partners, with separate access from ER 2

November 11th

Gemini Robotics 2.5 begins a partner rollout. What its task results show, how access differs from ER 2, and what deployment teams should verify.

Read more
New Models
Gemini 4 Carbon standard API input and output prices double from $2 and $10 to $4 and $20 per million tokens on December 1, 2026.

Gemini 4 Carbon opens to paying users, with API prices set to double in December

November 10th

Gemini 4 Carbon expands beyond early access. Compare three benchmark snapshots, public access routes and the December API price change.

Read more
New Models
November 9 benchmark comparison of Muse Ember, Claude Haiku 5.5 and GPT-6 Luna at maximum reasoning, with row-leading values bolded.

Muse Ember launches: what the benchmarks show for low-cost agent work

November 9th

Meta’s Muse Ember launches for agent workflows. Compare three benchmarks, launch API pricing and the long-prompt threshold before choosing a cheaper model.

Read more
New Models
Three sourced benchmark rows comparing Fable 5.2, Fable 5.1, Opus 5.5 and GPT-6 Astra, with each row's highest reported percentage in bold.

Claude Fable 5.2 launches with lower pricing and a separate Mythos access route

November 9th

Fable 5.2 is available for demanding technical work. Compare three benchmark snapshots, API pricing and the separate verification route for Mythos 5.2.

Read more
New Models
Tencent-reported xhigh replay throughput comparing autoregressive decoding, DFlash2 and DSpine across five workloads, in output tokens per second.

Tencent adds DSpine for Qwen3.8-27B, targeting faster self-hosted inference

November 8th

Tencent's new draft checkpoint accelerates Qwen3.8-27B inference in developer replay tests. Compare the results and what to measure before deployment.

Read more
New Models
Pruna-reported B200 bf16 timings: MiniMax-H3 versus P-H3-Image, 12.4 versus 2.6 seconds at 4 MP and 30.6 versus 5.8 seconds at 8 MP; lower is better.

Pruna releases P-H3-Image, with US deployment limited by its license

November 8th

P-H3-Image offers smaller image-model configurations, but US licensing and pipeline-specific memory requirements are central to deployment.

Read more
New Models
Original diagram of Gemini Omni 1.2 Flash’s draft, refine and upscale workflow, with access caveats and the official Gemini family logo.

Google releases Gemini Omni 1.2 Flash with longer video clips and lower pricing

November 5th

Gemini Omni 1.2 Flash adds 15-second video and lower output pricing. Access restrictions, token billing and evaluation limits matter before adoption.

Read more
New Models
1/3
1 2 3

Newsletter Subscription

Receive the latest AI news

Blog logo

PromptsZone

AI news and analysis from PromptsZone: model launches, research, AI tools, and the business and policy developments shaping artificial intelligence.

RSS feed

© PromptsZone, 2026.

Made with Blogmaker

×