Liquid AI releases d1.2 and d1.2-ultra, decision models that return probabilities with zero output tokens
Liquid AI released d1.2 and d1.2-ultra, decision models that read text, images and audio and return calibrated probabilities with zero output tokens. d1.2-ultra scores 75 on the Artificial Analysis Decision Index v2.1.
Liquid AI has released two decision models, d1.2 and d1.2-ultra. Both read text, images and audio and return calibrated probabilities with zero output tokens, according to MarkTechPost's report on the release. d1.2-ultra scores 75 on the Artificial Analysis Decision Index v2.1.
What was released
- d1.2 and d1.2-ultra: omnimodal decision models from Liquid AI.
- Inputs: text, images and audio.
- Output: calibrated probabilities, produced with zero output tokens.
- Benchmark: d1.2-ultra scores 75 on the Artificial Analysis Decision Index v2.1. MarkTechPost's headline says the models top the AA Decision Index.
Context window, parameter count, pricing and training cutoff date are not yet disclosed in the information available at publication. Pricing is not yet disclosed.
What this means
Most language models produce answers as generated text. The d1.2 models instead return a calibrated probability, and they do so with zero output tokens. For decision tasks, that means the output is a probability rather than a written answer.
The models accept text, images and audio, so a single model can handle decisions over all three input types. d1.2-ultra's score of 75 on the Artificial Analysis Decision Index v2.1 gives builders a reference point for this model class. Because the index is versioned (v2.1), scores should be compared only against results from the same version.
Related Articles
Liquid AI releases LiquidCodec, a 74M-parameter streaming audio codec running at 12.5 Hz with 12 codebooks
Liquid AI released LiquidCodec, a streaming audio codec of about 74M parameters. It runs at 12.5 Hz with 12 codebooks and encodes about 361x faster than real time on a single CPU thread. It powers LFM3-Audio-2B and LFM3-Omni-2B.
Huawei Cloud releases CloudRobo-WM1.2 world action model, claims first place on RoboTwin 2.0 and LIBERO-Pro 2.0
Huawei Cloud has released CloudRobo-WM1.2, an embodied world action model, and says it will be open source. The company claims first place on the RoboTwin 2.0 and LIBERO-Pro 2.0 robot manipulation benchmarks, with scores of 88.98% and 81.7%.
Tencent launches Hy4 flash at $0.07 per 1M input tokens, one-twentieth the per-task cost of Hy4
Tencent launched Hy4 flash on Monday, a lower-cost model for everyday work such as research, slide creation and spreadsheet editing. On Tencent's internal WorkIndex benchmark, it cost one-twentieth as much per task as the full Hy4 model. API list prices are $0.07 per million input tokens and $0.281 per million output tokens.
Qwen4 Max launches with 1M-token context and $1.50/$4.50 per 1M token pricing
Qwen4 Max, the flagship model in Alibaba's Qwen4 series, is listed on OpenRouter with a 1M-token context window. It accepts text, image and video input, returns text, and is priced at $1.50 per 1M input tokens and $4.50 per 1M output tokens.
Comments
Loading...