No release verdicts yet. Scored jobs show how many tools were tested. The rest are not tested.

Queue

CA6 · measured scores only

Continue vs Ollama

For Run a coding model locally or cheaply.

Ollama scored 81/100 and Continue scored 52/100. The gap is 29 points on the rubric. Nothing else is inferred.

Last tested:

Continue

52 / 100

Rank 3 on this job · tested 16 Oct 2026

Ollama

81 / 100

Rank 1 on this job · tested 15 Oct 2026

Rubric breakdown

  • Job result (40)12 Continue32 Ollama
  • Speed and effort (15)6 Continue12 Ollama
  • Price and free tier (15)15 Continue15 Ollama
  • Ease of start (10)6 Continue6 Ollama
  • Reliability and trust (10)6 Continue8 Ollama
  • Community signal (10)7 Continue8 Ollama

Evidence · Continue

Continue CLI 1.5.47 with a local Ollama config. Ping returned pong. The task reply in 86s included a trailing debounce and a handler whose import is an unterminated string. tsc reported TS1002. The repair reply was tool-call JSON and did not change the file. Evidence: evidence/2026-10-16/batch3/continue/.

Evidence · Ollama

Ollama 0.40.4, qwen2.5-coder:3b, 15Gi CPU, no GPU. Two prompts, 65.6s then 44.2s. Final files passed tsc and the debounce/REST checks. Evidence: evidence/2026-10-16/ca6/ollama/.

Back to Run a coding model locally or cheaply · How we rank

Newsletter

The weekly ranking note

One email when a job gets a tested top five, a price changes, or a tool ships something we re-test. No sponsored placements.