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

    Databricks rolls out Astra to all roughly 3,500 engineers

    Databricks' rollout announcement says Astra outperformed Opus 5 and Sol 5.6 on highly complex tasks in a roughly 200-user pilot. Engineers given Astra increased overall coding spend by around 60% versus baseline.

    Greg BrockmanGB
    Jonathan FrankleJF
    Matei ZahariaMZ
    12 Sources, ,

    TLDR

    Databricks announced on September 16, 2026, that it had rolled out Astra to all roughly 3,500 engineers following a pilot with around 200 users. The announcement reports stronger performance than Opus 5 and Sol 5.6 on highly complex tasks, especially high-level system design, but says meaningful improvement on medium- or low-complexity coding tasks remained unclear. Engineers given Astra increased overall coding spend by around 60% versus baseline. To encourage selective use, the announcement says engineers get an Astra-specific sub-budget, with lower-cost models preferred for everyday tasks. It also says Databricks lacked robust comparisons with Fable because data-retention policies had kept that model from being rolled out widely.

    Combined views

    1.3M

    12 Sources, first seen 21d ago

    Combined views

    1.3M

    12 Sources, first seen 21d ago

    4.8K likes
    21d ago
    first seen 21d ago
    4.8K likes262 comments1.3K saves255 reposts
    262 comments
    1.3K saves
    255 reposts

    Sentiment

    Positive71.4%28.6%Negative

    Summary

    Many accounts praised Databricks' full Astra rollout for strong real-world gains on hard tasks and clearer budget insights, while others called the model overpriced and objected to the 60% spending increase plus safety discussions.

    Based on 28 sentiment-bearing replies from 28 accounts across 3 conversations.

    Sentiment

    Positive71.4%28.6%Negative

    Summary

    Many accounts praised Databricks' full Astra rollout for strong real-world gains on hard tasks and clearer budget insights, while others called the model overpriced and objected to the 60% spending increase plus safety discussions.

    Based on 28 sentiment-bearing replies from 28 accounts across 3 conversations.

    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    12 Sources

    Patrick Wendell@pwendellToday we rolled out Astra to every engineer at Databricks (N=~3500). Some notes that may be helpful to others: 1. Astra unambiguously out performs our previous highest-end models (Opus 5, Sol 5.6) on highly complex tasks, especially those related to high level system design or long range horizontal tasks. 2. Engineers given Astra increased overall coding spend by around 60% compared to baseline. 3. It is not clear Astra meaningfully improves on medium/low complexity coding tasks compared to earlier models. We suspect those tasks are mostly saturated (i.e. perfectly executed) by existing models. 4. We learned above by piloting Astra with around 200 users to gain signal on both quality and cost. We use Unity AI Gateway to do cohort-based experiments for all new models. 5. We give engineers a sub-budget specific to Astra to encourage them to use Astra selectively on complex tasks while preferring lower cost models for everyday tasks. Our engineers are able to mix-and-match tools and models within their overall budget envelope (we also allow for increased budgets through various mechanisms). Note: We do not have robust comparisons of Astra-vs-Fable because we have net yet rolled out Fable widely due to data retention policies.21d
    Jonathan Frankle@jefrankleRT @pwendell: Today we rolled out Astra to every engineer at Databricks (N=~3500). Some notes that may be helpful to others: 1. Astra unam…21d
    Matei Zaharia@matei_zahariaRT @pwendell: Today we rolled out Astra to every engineer at Databricks (N=~3500). Some notes that may be helpful to others: 1. Astra unam…21d
    Yuchen Jin@Yuchenj_UWWe rolled out GPT-6 Astra to every Databricks engineer today. It beats Claude Opus 5 on the hardest, long-horizon tasks and increased our coding spend by 60%. I think it’s the strongest model. Expensive. But hopefully worth it.21d
    tae kim@firstadopterOpenAI is gaining market share due to its strong performance and Anthropic's onerous data retention policy. “Astra unambiguously out performs our previous highest-end models (Opus 5, Sol 5.6) on highly complex tasks” “Engineers given Astra increased overall coding spend by around 60% compared to baseline” “We do not have robust comparisons of Astra-vs-Fable because we have net yet rolled out Fable widely due to data retention policies”21d
    Rohan Paul@rohanpaul_ai3500 engineers just got Astra-pilled at Databricks.21d
    Greg Brockman@gdbgreat notes on wall-to-wall deployment of astra for engineers at databricks:21d
    Andy Konwinski@andykonwinskitldr: use astra for complex coding tasks and a cheaper model for less complex tasks "Engineers given Astra increased overall coding spend by around 60%"20d

    12 Sources

    Patrick Wendell@pwendellToday we rolled out Astra to every engineer at Databricks (N=~3500). Some notes that may be helpful to others: 1. Astra unambiguously out performs our previous highest-end models (Opus 5, Sol 5.6) on highly complex tasks, especially those related to high level system design or long range horizontal tasks. 2. Engineers given Astra increased overall coding spend by around 60% compared to baseline. 3. It is not clear Astra meaningfully improves on medium/low complexity coding tasks compared to earlier models. We suspect those tasks are mostly saturated (i.e. perfectly executed) by existing models. 4. We learned above by piloting Astra with around 200 users to gain signal on both quality and cost. We use Unity AI Gateway to do cohort-based experiments for all new models. 5. We give engineers a sub-budget specific to Astra to encourage them to use Astra selectively on complex tasks while preferring lower cost models for everyday tasks. Our engineers are able to mix-and-match tools and models within their overall budget envelope (we also allow for increased budgets through various mechanisms). Note: We do not have robust comparisons of Astra-vs-Fable because we have net yet rolled out Fable widely due to data retention policies.21d
    Jonathan Frankle@jefrankleRT @pwendell: Today we rolled out Astra to every engineer at Databricks (N=~3500). Some notes that may be helpful to others: 1. Astra unam…21d
    Matei Zaharia@matei_zahariaRT @pwendell: Today we rolled out Astra to every engineer at Databricks (N=~3500). Some notes that may be helpful to others: 1. Astra unam…21d
    Yuchen Jin@Yuchenj_UWWe rolled out GPT-6 Astra to every Databricks engineer today. It beats Claude Opus 5 on the hardest, long-horizon tasks and increased our coding spend by 60%. I think it’s the strongest model. Expensive. But hopefully worth it.21d
    tae kim@firstadopterOpenAI is gaining market share due to its strong performance and Anthropic's onerous data retention policy. “Astra unambiguously out performs our previous highest-end models (Opus 5, Sol 5.6) on highly complex tasks” “Engineers given Astra increased overall coding spend by around 60% compared to baseline” “We do not have robust comparisons of Astra-vs-Fable because we have net yet rolled out Fable widely due to data retention policies”21d
    Rohan Paul@rohanpaul_ai3500 engineers just got Astra-pilled at Databricks.21d
    Greg Brockman@gdbgreat notes on wall-to-wall deployment of astra for engineers at databricks:21d
    Andy Konwinski@andykonwinskitldr: use astra for complex coding tasks and a cheaper model for less complex tasks "Engineers given Astra increased overall coding spend by around 60%"20d