Google launches Gemini Decisions API and Gemini Decision 1 model for closed-set scoring
Article timeline
The articles in this story. Times are ET.
🚨 AI News | TestingCatalogXDaily AI brief covers Gemini Decision 1, Grok Bot 1.0, and an NVIDIA CEO reportAIGoogle's Gemini Decision 1 is live in the Gemini API and AI Studio, scoring up to 1,000 questions per call at $0.10 per 1M input tokens with no output cost. xAI's Grok Bot 1.0 adds shareable bots, Slack and Teams collaboration, and meeting invites, and SpaceX ordered 100 GW/year of HJT solar manufacturing equipment from Singulus. Bloomberg reports NVIDIA's board is preparing to oust CEO Jensen Huang, sending shares sharply lower after hours.
- Latent SpaceBlogPickLatent Space's daily roundup covers Gemini Decision 1 and a wide range of AI model and research news
AILatent Space's daily AINews roundup covers Google's Gemini Decision 1 launch, which Sundar Pichai and Logan Kilpatrick announced as a calibrated decision model priced at $0.10 per million input tokens. The newsletter also summarizes frontier model evaluations, open-weight releases, robotics and science work, safety research, and industry news in a single digest.
- SiliconANGLE · AINewsGoogle launches Gemini Decisions API with its Gemini Decision 1 model
AIGoogle has launched the Gemini Decisions API, powered by a new model called Gemini Decision 1 that answers only yes-no, multiple-choice, and rating questions. Google says the model is nearly five times cheaper than its entry-level Gemini 3.5 Flash-Lite, charges $0.10 per 1 million input tokens, and does not charge for output tokens. Google says it beat OpenAI's Decisions API on seven of nine benchmarks, while Jev was slightly more cost-efficient.
meng shaoXGoogle launches Gemini Decision 1, a scoring model that returns no textAIGoogle introduces Gemini Decision 1 in the Gemini API, a decision model that scores up to 1,000 independent questions in one inference pass. It returns calibrated probabilities and confidence values, with zero output tokens. Input costs $0.10 per 1M tokens with no output charge, and the free tier is limited to 200 requests per minute, 50K tokens, and 10K requests per day.
- Simon WillisonBlogGoogle launches Gemini Decision 1, a decision model API priced at 10 cents per million input tokens
AIGoogle launches Gemini Decision 1, a decision model API that answers yes/no, choice, and rubric-scoring questions across text, images, audio, video, and PDFs. It is priced at 10 cents per million input tokens, the same as the GPT-6 Luna decision model, with output free. Simon Willison reports that the model returns probabilities such as 0.92 for a robot in his test video, where Jev and OpenAI return 1.0 or 0.0.
- MarkTechPostNewsGoogle launches Gemini Decisions API with Gemini Decision 1 at $0.10 per 1M input tokens
AIGoogle has launched the Gemini Decisions API, powered by the new model Gemini Decision 1 (gemini-decision-1). The API scores up to 1,000 typed questions against one shared input and returns calibrated probabilities with zero output tokens billed. Google says it runs in 150-250 ms end to end and is available through the Gemini Developer API and a no-cost tier in Google AI Studio.
Josh WoodwardOfficialGemini Decision 1 launches in the Gemini API at $0.10 per 1M input tokensAIJosh Woodward says the Gemini team builds with the product regularly and that it makes their apps faster and smarter. Google introduces Gemini Decision 1, a frontier decision model in the Gemini API, priced at $0.10 per 1M input tokens with no additional cost. It natively supports image, audio, video, and PDF inputs, has a 256K context window, and can be tried free in the API.
SantiagoXSantiago says Jev is undercutting Google's Gemini Decision 1 on priceAISantiago (@svpino) says Jev's model costs $0.20 less per 1M tokens than Google's Gemini Decision 1, which Google's Logan Kilpatrick, quoted in the post, lists at $0.10 per 1M input tokens. Santiago says he is surprised Google cannot compete on price.
- The DecoderNewsGoogle releases Gemini Decision 1, which scores 1,000 questions in a single pass
AIGoogle DeepMind introduces the Gemini Decisions API and its underlying model, Gemini Decision 1, which processes up to 1,000 independent yes/no, multiple-choice, and scale-rating questions in one pass across text, images, audio, video, and PDFs. It returns probabilities and confidence values in 150 to 250 milliseconds. Google says the API costs $0.10 per million input tokens with no output-token charge, about five times cheaper than Gemini 3.5 Flash-Lite, and is available now through Google AI Studio.
👩💻 Paige BaileyOfficialGoogle launches Gemini Decisions API for fast classification and routingAIGoogle introduces the Gemini Decisions API, powered by the new Gemini Decision 1 model, for routing, tool selection, safety checks and grading without full text generation. The API evaluates up to 1,000 questions per request, costs $0.10 per 1M input tokens and $0 for output tokens, and accepts one shared input across text, images, audio, video and PDFs. Google's Gemini account says developers have used its Flash-Lite and smaller Flash models as classifiers and calls the new option tinier, faster and more cost-effective.
- 👩💻 Paige BaileyOfficialGoogle introduces Gemini Decision 1 with $0.10 per million input tokens
AIGoogle introduces Gemini Decision 1, a frontier decision model available in the Gemini API. It costs $0.10 per 1M input tokens with no additional charge, supports image, audio, video, and PDF input, and has a 256K context window. Paige Bailey says the team is making smaller, efficient classification models available through the API and asks developers to test it.
elvisXGoogle introduces Gemini Decision 1, a decision model in the Gemini APIAIGoogle introduces Gemini Decision 1, a frontier decision model in the Gemini API that accepts text, images, audio, video, and PDFs. It has a 256K context window and costs $0.10 per 1M input tokens, with free API trials offered.
Google AI StudioOfficialPickGoogle launches the Gemini Decisions API for fast, calibrated multimodal decisionsAIGoogle launches the Gemini Decisions API, powered by the Gemini Decision 1 model (gemini-decision-1), in Google AI Studio. The endpoint scores up to 1,000 yes/no, choice, and rubric questions in a single pass across text, images, audio, video, and PDFs, returning probabilities and confidence scores at $0.10 per 1M input tokens with no output tokens billed. Google says Gemini Notebook, Stitch, CC, and Search already use it for routing and safety checks.
Andrew CurranXGoogle launches Gemini Decisions API for closed-set choice scoringAIGoogle launches the Gemini Decisions API, powered by its Gemini Decision 1 model, in Google AI Studio. It is built for scoring up to 1,000 closed-set questions across text, images, audio, video, and PDFs in one pass, priced at $0.10 per 1M input tokens with no output-token charge. The image shows Gemini Decision 1 scoring higher than Gemini 3.5 Flash-Lite on several benchmarks, with lower calibration error.
- 🚨 AI News | TestingCatalogXGoogle launches Gemini Decisions API powered by Gemini Decision 1
AIGoogle launches the Gemini Decisions API, powered by its new Gemini Decision 1 model, in Google AI Studio. The API is built for closed-set choices, scoring up to 1,000 questions across text, images, audio, video and PDFs in a single pass. It charges $0.10 per 1M input tokens with no output token charges, according to the post.
GoogleOfficialGoogle shows Gemini Decision 1 assessing delivery photos with a yes/no checklistAIGoogle says a checklist of yes/no, multiple-choice, and rating questions tells Gemini what to check in a photo and what action to take. Gemini Decision 1 scans 100 delivery photos in 4.5 seconds, nearly 4x faster than Gemini 3.5 Flash-Lite, answering 16 questions per photo at once. Questions include whether the box is damaged, whether it sits under a porch, a 0 to 3 delivery quality rating, and routing to a claim review.
- GoogleOfficialGoogle introduces Gemini Decisions API in Google AI Studio
AIGoogle introduces the Gemini Decisions API in Google AI Studio, powered by its new Gemini Decision 1 model. Instead of generating text token by token, the model evaluates input across text, images, audio, video, and PDFs against up to 1,000 questions in a single pass, with zero output tokens billed.
Google for DevelopersOfficialPickGoogle introduces the Gemini Decisions API for routing and gradingAIGoogle introduces the Gemini Decisions API, powered by its new Gemini Decision 1 model, for tool selection, safety checks and grading in apps and agents. The API evaluates up to 1,000 questions per request and costs $0.10 per 1M input tokens with $0 for output tokens. It accepts one shared input across text, images, audio, video and PDFs, and developers can start by getting an API key in Google AI Studio.
Vercel DevelopersOfficialGemini Decision 1 becomes available on Vercel AI GatewayAIVercel says Google's Gemini Decision 1 is now available on AI Gateway. The model returns structured decisions with probabilities from text, images, documents, video, and audio inputs, for classification, routing, and scoring.
- Google AI StudioOfficialGoogle AI Studio shares three apps built with the Decisions API
AIDoorstep Parcels, Speed Sketch, and Find the Zookeeper. Each app is available to try in AI Studio through a linked page.
OpenRouterOfficialGoogle releases Gemini Decision 1, a decision model available on OpenRouterAIOpenRouter says Google has released Gemini Decision 1, its first decision model, available on OpenRouter from launch. The model answers yes/no, multiple-choice, and scoring questions with a probability, across text, images, audio, video, and PDFs. It costs $0.10 per million input tokens, with free output, and has a 262K context window.
Artificial AnalysisOfficialArtificial Analysis publishes Gemini Decision 1 results on its Decision IndexAIArtificial Analysis says full results for Gemini Decision 1 are now available on its Decision Index evaluation page. The post links to the evaluation but gives no scores or comparisons.
- Artificial AnalysisOfficialArtificial Analysis reports Gemini Decision 1 at 211ms per task
AIArtificial Analysis reports that Gemini Decision 1 takes 211ms per task on the Decision Index, faster than Drex Max v1.1 (283ms) and GLiDE v1.2 (284ms). It is slower than Celeris-1.1 Decision (61ms), Elephant 9B (79ms) and Jev 1.14 (92ms).
- Artificial AnalysisOfficialGemini Decision 1 leads decision models on Artificial Analysis knowledge and reasoning tests
AIArtificial Analysis says Gemini Decision 1 scores highest among decision models on the knowledge and reasoning evaluations in its Decision Index. It scores 66 on AA-Omniscience-Decision, level with GPT-6 Luna (Max) and ahead of d1-ultra (64), answering 45% of questions correctly. On HLE-MultiChoice it scores 49, leading on Decision Score through a Brier score of 0.20, among the lowest on that evaluation.
- Artificial AnalysisOfficialGemini Decision 1 costs $345 per 1M tasks on Artificial Analysis index
AIArtificial Analysis reports that Gemini Decision 1 costs $345 per 1M tasks across its Decision Index, close to GPT-6 Luna at $339. The higher cost comes from token pricing, with Gemini Decision 1 at $0.10 per 1M input tokens versus $0.03 for Drex Max v1.1, which costs $103 per 1M tasks.
- Artificial AnalysisOfficialGoogle's Gemini Decision 1 scores 66 on Artificial Analysis Decision Index
AIArtificial Analysis reports that Google's Gemini Decision 1 scores 66 on its Decision Index v2.1, seven points behind Drex Max v1.1 and GLiDE v1.2, which both score 73. Gemini Decision 1 costs $345 per 1M Tasks, more than three times the $103 for Drex Max v1.1, because its input token price is $0.10 per 1M versus $0.03. It leads knowledge and reasoning benchmarks but trails on AA-Paperclip, answering all 15 questions correctly on 33% of documents versus 63% for Drex Max v1.1.
Logan KilpatrickOfficialGoogle introduces Gemini Decision 1 model in the Gemini APIAIGoogle introduces Gemini Decision 1, a frontier decision model available in the Gemini API. It costs $0.10 per 1M input tokens with no additional cost, and supports image, audio, video, and PDF input with a 256K context window. Developers can try it for free in the API.
- Google AI StudioOfficialGoogle launches Gemini Decisions API powered by Gemini Decision 1
AIGoogle introduces the Gemini Decisions API, powered by Gemini Decision 1, its new multimodal decision model. The API returns fast, calibrated decisions on up to 1,000 questions at once across text, images, audio, video, and PDFs. Developers can try it today in AI Studio and through the Gemini API.
Philipp SchmidOfficialPickGoogle ships Gemini Decision 1 and a new API six weeks after internal prototypeAIGoogle says Gemini Decision 1 and a new API are shipping together, six weeks after a prototype began inside Google DeepMind and was tested across Google teams. Philipp Schmid, who says he helped design, build, and evaluate the API, credits a small team of research, engineering, and product staff. Docs, a blog post, and a playground are linked for developers to try it.
- Philipp SchmidOfficialGoogle launches Gemini Decisions API with Gemini Decision 1 model
AIGoogle announces the Gemini Decisions API for yes/no, multiple choice, and rubric questions, powered by its new Gemini Decision 1 model. The API accepts text, images, audio, video, and PDF input, up to 1,000 questions per request with a 256k context window, and returns a confidence score on every answer. It is priced at $0.10 per 1M input tokens with no output charge, with 150-250 ms end-to-end latency.
Sundar PichaiOfficialPickGoogle launches Gemini Decisions API with new Gemini Decision 1 modelAIGoogle launches the Gemini Decisions API, powered by its new Gemini Decision 1 model in Google AI Studio. Instead of generating text token by token, it scores up to 1,000 closed-set questions across text, images, audio, video, and PDFs in one pass. It supports yes/no, multiple choice with up to 255 options, and rubric scoring, billed at $0.10 per 1M input tokens with no output-token charge, which Google says is about 5x cheaper than Flash-Lite.
- Google · AI blogOfficialPickGoogle launches Gemini Decisions API for fast multimodal decisions
AIGoogle introduces the Gemini Decisions API, powered by the Gemini Decision 1 model, for fast closed-set decisions across text, images, audio, video, and PDFs. The model is priced at $0.10 per 1M input tokens with zero output tokens billed, and it is available through the Gemini API in Google AI Studio.
- 🚨 AI News | TestingCatalogXGoogle's Gemini Decision 1 model reportedly spotted on Google APIs
AITesting Catalog reports that a Gemini model named Decision 1 has been spotted on Google APIs. The post says no Flash variant appeared, and it asks whether the sighting is true.
- Gemini API ChangelogOfficialPickGoogle releases Gemini Decision and the Decisions API for classification tasks
AIGoogle releases Gemini Decision (gemini-decision-1) and the Decisions API. The source says they handle fast, calibrated binary, categorical, and rubric decisions across text, images, audio, video, and PDFs with zero output tokens.