chini-003-twitter-timeline

Social Timeline (Twitter-style fanout)

Generate a personalized timeline for millions of users. Don't melt when a celebrity posts.

Source: Classic system-design interview corpus (Twitter / Instagram timeline)

Prompt

Design the timeline-generation system for a social network.

Functional:
- POST /tweet appends a post by a user.
- GET /timeline returns the most recent posts from accounts a user follows.
- Reads vastly outnumber writes (~100:1).

Non-functional:
- A celebrity post (high-fanout write) must NOT degrade timeline reads for other users.
- Survive a 10x read spike (a viral event).
- A timeline-cache failure should degrade gracefully (slower reads), not drop posts.

Return a Chinilla CanvasState. You will likely need a write path, a read path, a cache, and some kind of fanout/queue between them. Include a rate-limit behavior on the write side.

Constraints

Max components: 14
Required behaviors: queue, ratelimit, storage
Monthly budget: $1500

Stress scenarios

Baseline reads + writes

baseline

Mixed traffic at normal rates.

10x viral spike

spike

Read traffic explodes 10x. Reads must survive even if write path slows.

Timeline cache failure

outage

Cache layer drops out. Reads must fall back, not error.

Noisy network

cascade

Process-time jitter at 25%. Tests resilience to flaky internal calls.

Pass criteria (overall)

Min stability score: 70
Max drop rate: 5.0%
Min delivery rate: 90.0%
Max errors: 5

Submit your run

Submissions go through the chini-bench CLI. It calls your model with your key, scores the result locally, and posts to the leaderboard. Nothing leaves your machine except the canvas it produces.

End-to-end:

pip install git+https://github.com/collapseindex/chini-bench-cli.git
export OPENROUTER_API_KEY=...

chini-bench run chini-003-twitter-timeline \
  --provider openrouter --model google/gemini-2.0-flash-001 \
  --as alice --x alice --linkedin alice-builds

Or inspect the prompt first:

chini-bench prompt chini-003-twitter-timeline

Providers: openai · anthropic · google · openrouter · ollama

Leaderboard

Rank	Submitter	Model	Score	Stability	Delivery	Design	Pass	Links
#1	alex default	G google/gemini-3.1-pro-preview	79	57.0	100.0	100.0	✗	X
#2	alex default	O openai/gpt-5.4	78	60.0	100.0	100.0	✗	X
#3	alex default	A anthropic/claude-sonnet-4.6	76	52.0	100.0	100.0	✗	X
#4	alex default	X x-ai/grok-4.20	42	16.0	47.0	100.0	✗	X

Per-scenario breakdown of the top run

Scenario	Health	Drop rate	Delivered	Pass
baseline	68.0	0.4%	474	✗
viral-spike	62.0	1.6%	4103	✓
cache-failure	60.0	0.0%	60	✗
fanout-jitter	38.0	9.5%	85	✗

How is this scored? →