Last updated 16 Nov 2026. Sources last checked 16 Nov 2026.
A sourced tracker of AI decision models
Decision models return a choice, a yes-or-no probability, or a score — not a written answer. This catalog lists publicly announced products, with a source link on every row. We do not invent entries, and we do not run a hidden ranking. Other tabs on this site cover blockchains, AI agents, tokenized real-world assets, and humanoid robots, also from official public sources only.
What is a decision model? · 25 sourced models · Timeline · Compare two entries · RSS feed
What changed this week
- d1.2-ultra · 16 Nov 2026
- d1.2 · 16 Nov 2026
- Gemini Decision 1 · 11 Nov 2026
The catalog
| What it is for | Access and pricing | Source | |||
|---|---|---|---|---|---|
| d1.2-ultra | Liquid AI | 16 Nov 2026 | The higher-capacity d1.2 model, with the same new task types and audio input as d1.2. | As stated on 16 Nov 2026: $0.042 per million input tokens; input tokens only. | Official announcement |
| d1.2 | Liquid AI | 16 Nov 2026 | Updated d1 decision model with audio input and five new task types: reading a value (text, number, or date), finding spans, locating objects in a picture,… | As stated on 16 Nov 2026: $0.02 per million input tokens; input tokens only. | Official announcement |
| Gemini Decision 1 | Google DeepMind | 11 Nov 2026 | The model behind the Gemini Decisions API: a new endpoint for fast, closed-set, multimodal decisions. | As stated on 11 Nov 2026: $0.10 per million input tokens and $0 for output tokens. The shared input is billed once per request; each question adds only its own instruction and criteria tokens. | Official announcement |
| Koa Decisions | Salesforce | 6 Nov 2026 | The next version of Salesforce's HyperClassifier decision model, used inside Agentforce to route requests and make single-step choices, yes/no decisions, or… | Salesforce said pricing and packaging are subject to change, that availability may vary by region, and that purchasing decisions should be based on products currently available under customer agreements. No public list price was stated in the announcement. | Official announcement |
| Hume Audio Decision API | Hume AI | 2 Nov 2026 | An audio-native decision API that answers developer-defined questions about speech and sound. | A public list price was not stated in the 2 Nov 2026 announcement. | Official announcement |
| Parallel Decisions API | Parallel | 25 Oct 2026 | A web-grounded decision engine: it answers closed-ended questions using current web evidence and returns typed answers, probabilities, and citations. | As stated in the launch post: $1 per 1,000 calls, with 5,000 free requests to get started. Batching multiple questions about a shared context is billed as a single call. | Official announcement |
| Zoom Decisions Pro | Zoom | 19 Oct 2026 | The higher-capacity Zoom Decisions model, sharing one interface with Zoom Decisions and aimed at more demanding judgments such as checking contract clauses,… | As stated at launch on 19 Oct 2026: $0.10 per million input tokens. | Official announcement |
| Zoom Decisions | Zoom | 19 Oct 2026 | A decision API in Zoom AI Services for high-volume classification, routing, triage, and tagging. | As stated at launch on 19 Oct 2026: $0.042 per million input tokens. | Official announcement |
| d1-ultra | Liquid AI | 15 Oct 2026 | A multimodal decision model in the d1 family, described by Liquid AI as aimed at harder, messier real-world use cases such as spotting a defect against a spec… | As stated in the 15 Oct 2026 announcement: $0.042 per million input tokens. | Official announcement |
| Strands Decider E4B | Strands Agents | 15 Oct 2026 | A Gemma 4-based open decision model in the Strands Decider family (E4B size). Same noul, choice, and score interface as the other sizes in this release. | Open source / self-hosted. No hosted list price was stated in the announcement. | Official announcement |
| Strands Decider E2B | Strands Agents | 15 Oct 2026 | A Gemma 4-based open decision model in the Strands Decider family (E2B size). Accepts the same noul, choice, and score question types as Strands Decider 2B, so… | Open source / self-hosted. No hosted list price was stated in the announcement. | Official announcement |
| Strands Decider 26B-A4B | Strands Agents | 15 Oct 2026 | A Gemma 4 mixture-of-experts open decision model in the Strands Decider family. The announcement said the 26B-A4B model activates about 4B of its 26B… | Open source / self-hosted. No hosted list price was stated in the announcement. | Official announcement |
| Strands Decider 12B | Strands Agents | 15 Oct 2026 | A Gemma 4-based open decision model in the Strands Decider family (12B size). Same noul, choice, and score interface as the other sizes in this release. | Open source / self-hosted. No hosted list price was stated in the announcement. | Official announcement |
| Granite Switch 4.1 | IBM | 15 Oct 2026 | Modular Granite Switch models (8B and 30B) that combine a general-purpose Decision aLoRA, a specialized Guardian safety-decision adapter, and speech-to-text in… | Released as compose-it-yourself open models; no public hosted list price was stated in the 15 Oct 2026 announcement. | Official announcement |
| Microsoft-Decision-1 | Microsoft | 9 Oct 2026 | A decision-scoring model for routing, classification, prioritization, verification, and workflow control. | As stated in the 9 Oct 2026 announcement: $0.042 per million input tokens; output tokens free. | Official announcement |
| Clef-omni | Cloudflare | 9 Oct 2026 | Multimodal addition to the Clef family that accepts text, images, audio, and video in one request and returns schema-bound, calibrated decisions. | As stated in the 9 Oct 2026 announcement: $0.15 per million input tokens. See Cloudflare's developer docs for current rates, including how media is converted to input tokens. | Official announcement |
| Snowflake Decision | Snowflake | 8 Oct 2026 | A decision model in Cortex AI Functions, called through AI_COMPLETE, for high-volume classification, scoring, filtering, routing, and triage over tables in… | No public list price was stated in the announcement. Access was described as private preview. | Official announcement |
| Open d1 | Liquid AI | 7 Oct 2026 | Open-weight models in the d1 family for edge and local use: d1-3B (text and images) and d1-omni-600M (text plus image, or text plus audio). | Open weights; self-hosted. No hosted list price was stated in the announcement. | Official announcement |
| H2O-Lightning-4B | H2O.ai | 7 Oct 2026 | A 4-billion-parameter open decision model, built on Qwen3.5-4B, that returns a probability for every bounded answer in a single forward pass. | Open weights; self-hosted. H2O.ai cited a JevBench board estimate of about $0.021 per 1,000 decisions for the self-hosted v1.1 row, marked as an estimate because open weights have no public price list. | Official announcement |
| Decisions API (GPT-6 Luna) | OpenAI | 6 Oct 2026 | A dedicated Decisions API that evaluates text, images, and structured inputs and returns a probability, a choice from a fixed set, or a rubric score. | As stated in the 22 Oct 2026 general-availability post: $0.10 per million input tokens; cached input $0.01 per million tokens; no output-token charges. See OpenAI's pricing guide for cache-write rates and regional options. | Official announcement |
| d1 | Liquid AI | 5 Oct 2026 | Liquid AI's first decision model with vision. It reads text, images, or both plus one or more questions in a single forward pass and returns probabilities… | As stated in the 5 Oct 2026 announcement: billed on input tokens only, at $0.04 per million input tokens; no output tokens. | Official announcement |
| Strands Decider 2B | Strands Agents | 1 Oct 2026 | A small, open-source 2-billion-parameter decision model for fast experimentation and local development. | Open source; self-hosted. No hosted list price was stated in the announcement. | Official announcement |
| Clef-flash | Cloudflare | 1 Oct 2026 | Latency-focused sibling of Clef for faster, lower-cost decisions. Open-weight, Jev-API compatible, and hosted on Workers AI. | As stated in Cloudflare's 9 Oct 2026 post: $0.038 per million input tokens, down from the earlier $0.09 rate. The hosted context window was reduced to 24k tokens as part of that price cut; Hugging Face weights were trained to support a larger window for self-hosting. See Cloudflare's developer docs for current rates. | Official announcement |
| Clef | Cloudflare | 1 Oct 2026 | Open-weight decision model for fast, typed classifications and routing in agent workflows. Hosted on Workers AI and released on Hugging Face under Apache 2.0. | As stated in Cloudflare's 9 Oct 2026 follow-up post: $0.24 per million input tokens. See Cloudflare's developer docs for current rates. | Official announcement |
| Jev | TypeSafe AI | 15 Sep 2026 | TypeSafe's first System One model: a frontier model built to return typed, calibrated decisions that software can use directly, rather than generated text. | As stated on 15 Sep 2026: $0.042 per million input tokens; output tokens free. | Official announcement |
No models match those filters.