• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI
    Announcement

    Perplexity releases open-weights pplx-decider-v1.1-27b

    The company says the multimodal model tops Decision Index 0.3 and costs half as much as v1.

    Aravind SrinivasAS
    Connor ShortenCS
    PerplexityPE
    10 Sources, ,

    TLDR

    Perplexity has released pplx-decider-v1.1-27b, which it describes as an updated open-weights multimodal decision model. The company says it scores highest on Hugging Face's private Decision Index 0.3 benchmark and costs $0.02 per million input tokens, half the price of v1. Perplexity's CTO also said a new checkpoint is available through the API and has been open-sourced.

    Combined views

    45.1K

    10 Sources, first seen 1h ago

    Combined views

    45.1K

    10 Sources, first seen 1h ago

    504 likes
    1h ago
    first seen 1h ago
    504 likes
    50 comments
    146 saves
    60 reposts
    50 comments
    146 saves
    60 reposts

    An updated decision model

    Perplexity has made pplx-decider-v1.1-27b available, describing it as an updated open-weights multimodal decision model.

    Featured Source

    The company's developer account says the model scores highest on Hugging Face's Decision Index 0.3 benchmark. Perplexity's CTO described it as a private benchmark. Perplexity's posts do not provide the benchmark methodology, so the ranking remains a company claim rather than an independently verified result.

    Price and availability

    Perplexity says v1.1 costs $0.02 per million input tokens, half as much as v1. The CTO presented the same rate as $20 per 1 billion tokens and said the company optimized inference.

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Today's Rank

    #5

    Today's Rank

    #5

    The CTO also said a new checkpoint is available through the API and has been open-sourced. Perplexity's posts do not specify a license or describe which materials are included beyond the checkpoint.

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    10 Sources

    Aravind Srinivas@AravSrinivasPerplexity Decision Model (open weights) scores the best on @huggingface Decision Index1h
    Perplexity Developers@perplexitydevspplx-decider-v1.1-27b, our updated open weights multimodal decision model, is now available. It scores the highest on the new @huggingface Decision Index 0.3 benchmark. pplx-decider-v1.1-27b costs half as much as v1, at $0.02 per million input tokens. https://huggingface.co/spaces/multimodalart/jev-decision-index1h
    Perplexity@perplexity_aiRT @perplexitydevs: pplx-decider-v1.1-27b, our updated open weights multimodal decision model, is now available. It scores the highest on…1h
    Antoine Chaffin@antoine_chaffinI wanted to brag about how we went from 56.4 to around 61.5 (with the same backbone!) on Decision Index v0.2 (3.5 points above the n°1, Jev) But in the meantime, Decision Index v0.3 got out… So, happy to brag about how we are already top 1 on this new leaderboard!1h
    Igor Carron@IgorCarronOn se pose des questions existentielles en France mais quand on a un français qui change la face du monde, on entends que des crickets. Meilleur que jev.1h
    Denis Yarats@denisyaratsa new checkpoint is available in the API and also open-sourced! it scores the best on the new @huggingface Decision Index 0.3 private benchmark. we also optimized inference and cut the cost by 2x: 1B tokens now costs only $20.1h
    Connor Shorten@CShorten30RT @antoine_chaffin: I wanted to brag about how we went from 56.4 to around 61.5 (with the same backbone!) on Decision Index v0.2 (3.5 poin…24m

    10 Sources

    Aravind Srinivas@AravSrinivasPerplexity Decision Model (open weights) scores the best on @huggingface Decision Index1h
    Perplexity Developers@perplexitydevspplx-decider-v1.1-27b, our updated open weights multimodal decision model, is now available. It scores the highest on the new @huggingface Decision Index 0.3 benchmark. pplx-decider-v1.1-27b costs half as much as v1, at $0.02 per million input tokens. https://huggingface.co/spaces/multimodalart/jev-decision-index1h
    Perplexity@perplexity_aiRT @perplexitydevs: pplx-decider-v1.1-27b, our updated open weights multimodal decision model, is now available. It scores the highest on…1h
    Antoine Chaffin@antoine_chaffinI wanted to brag about how we went from 56.4 to around 61.5 (with the same backbone!) on Decision Index v0.2 (3.5 points above the n°1, Jev) But in the meantime, Decision Index v0.3 got out… So, happy to brag about how we are already top 1 on this new leaderboard!1h
    Igor Carron@IgorCarronOn se pose des questions existentielles en France mais quand on a un français qui change la face du monde, on entends que des crickets. Meilleur que jev.1h
    Denis Yarats@denisyaratsa new checkpoint is available in the API and also open-sourced! it scores the best on the new @huggingface Decision Index 0.3 private benchmark. we also optimized inference and cut the cost by 2x: 1B tokens now costs only $20.1h
    Connor Shorten@CShorten30RT @antoine_chaffin: I wanted to brag about how we went from 56.4 to around 61.5 (with the same backbone!) on Decision Index v0.2 (3.5 poin…24m