Working Notes on Agent Systems/Brad Zhang

@teach_fireworks / X longform

Open AI Former CTO Mira's startup-general open source model Inkling Inkling: 975B, 41B total activations

Open AI Former CTO Mira's startup-general open source model Inkling Inkling: 975B, 41B total activations GLM-5.2: 744B, 40B total activations Supports 1M context...

July 15, 2026 路 1 min read

Open AI Former CTO Mira's startup-general open source model Inkling Inkling: 975B, 41B total activations GLM-5.2: 744B, 40B total activations Supports 1M context and adjustable inference budget. Coding capabilities are not as good as GLM 5.2 Ultra Large MoE, but Active Parameters are controlled very low. The knowledge capacity continues to expand, but the cost of reasoning does not rise. This is very critical for long-running agents. Thinking Effort is adjustable. Different inference budgets are allocated to different tasks, and model capabilities begin to become a schedulable resource. Token Efficiency In the future, the competition will not only be accuracy, but also Quality/Token and Quality/Cost. Native multimodality + open weighting. Enterprises can fine-tune, deploy, not just call APIs.

Visual summary

Article argument map

Generated from the post's content graph

FORMATTOPICCAPABILITYMARKETcoverscoverscoverscoverssignalssignalssignalssignalsFORMATlongform noteTOPICagentsTOPIClong running agentsTOPICmemoryTOPICinferenceCAPABILITYagent workflowCAPABILITYlong-running agentsCAPABILITYevaluationMARKETAI startupMARKETopen-source builders
Mermaid outline
flowchart LR
  format-long_post["longform note"]
  topic-agents["agents"]
  topic-long-running-agents["long running agents"]
  topic-memory["memory"]
  topic-inference["inference"]
  capability-agent-workflow["agent workflow"]
  capability-long-running-agents["long-running agents"]
  capability-evaluation["evaluation"]
  market-ai-startup["AI startup"]
  market-open-source-builders["open-source builders"]
  format-long_post -->|covers| topic-agents
  format-long_post -->|covers| topic-long-running-agents
  format-long_post -->|covers| topic-memory
  format-long_post -->|covers| topic-inference
  format-long_post -->|signals| capability-agent-workflow
  format-long_post -->|signals| capability-long-running-agents
  format-long_post -->|signals| capability-evaluation
  format-long_post -->|signals| market-ai-startup

Visual structure

Essay structure map

Built from summary and key paragraph positions

Open AI Former CTO Mira's startup-general open source model Inkling Inkling: 975B, 41...THESISOpen AI Former CTOMira's startup-generalopen source modelInkling Inkling: 975B,SIGNALOpen AI Former CTOMira's startup-generalopen source modelInkling Inkling: 975B,OPERATOROpen AI Former CTOMira's startup-generalopen source modelInkling Inkling: 975B,IMPLICATIONOpen AI Former CTOMira's startup-generalopen source modelInkling Inkling: 975B,
Mermaid outline
flowchart LR
  thesis["Open AI Former CTO Mira's startup-general open source model Inkling Inkling: 975B, 41B total activations GL..."]
  signal["Open AI Former CTO Mira's startup-general open source model Inkling Inkling: 975B, 41B total activations GL..."]
  operator["Open AI Former CTO Mira's startup-general open source model Inkling Inkling: 975B, 41B total activations GL..."]
  implication["Open AI Former CTO Mira's startup-general open source model Inkling Inkling: 975B, 41B total activations GL..."]
  thesis -->|frames| signal
  signal -->|develops| operator
  operator -->|lands in| implication

Source: View the original post