Microsoft lists 2 models and Scaledown 2. Scaledown is cheaper on a typical model ($0.03 per 1M tokens median). Below: line-up, price tiers, the cheapest model for each capability, and what each costs for five real workloads.
| Microsoft | Scaledown | |
|---|---|---|
| Models listed | 2 | 2 |
| Models with published pricing | 2 | 2 |
| Typical price (median, 3:1 blend per 1M) | $0.35 | $0.03 |
| Cheapest paid model | Microsoft: Phi 4 ($0.09) | ScaleDown: Extract ($0.03) |
| Longest context | WizardLM-2 8x22B (66K) | ScaleDown: Extract (1M) |
| Free models | 0 | 0 |
| Newest model | Microsoft: Phi 4 (2025-01-10) | ScaleDown: Extract (2026-10-13) |
| Price changes in the last 90 days | 1 | 0 |
For each workload: the cheapest model from each provider that can actually do it (fits the context window; tool calling required for the agent), and its monthly cost at list price.
| Workload (per month) | Microsoft | Scaledown |
|---|---|---|
| Customer-support chatbot 50,000 conversations — each a 6-turn conversation | $8.05 with Microsoft: Phi 4 | $3.00 with ScaleDown: Extract |
| RAG search over documents 20,000 questions — each a question answered from ~8 retrieved passages | $9.80 with Microsoft: Phi 4 | $4.80 with ScaleDown: Extract |
| Coding / tool-using agent 2,000 tasks — each a multi-step task with tool calls and file context | no suitable model | no suitable model |
| Batch summarisation 10,000 documents — each a 6,000-word report summarised to a page | $6.44 with Microsoft: Phi 4 | $3.20 with ScaleDown: Extract |
| Long-document analysis 500 documents — each a 150-page contract or codebase read in one pass | no suitable model | $2.00 with ScaleDown: Extract |
| You need | Microsoft | Scaledown |
|---|---|---|
| Any model | Microsoft: Phi 4 ($0.09) | ScaleDown: Extract ($0.03) |
| With tool calling | none listed | none listed |
| With image input | none listed | ScaleDown: Extract ($0.03) |
| With reasoning mode | none listed | none listed |
| 128K+ context | none listed | ScaleDown: Extract ($0.03) |
| 1M+ context | none listed | ScaleDown: Extract ($0.03) |
| Model type | Microsoft | Scaledown |
|---|---|---|
| text+image+file->text | 0 | 2 |
| text->text | 2 | 0 |
| Price per 1M (3:1 blend) | Microsoft | Scaledown |
|---|---|---|
| Free | 0 | 0 |
| Under $0.50 | 1 | 2 |
| $0.50 – $3 | 1 | 0 |
| $3 – $10 | 0 | 0 |
| $10 and up | 0 | 0 |
Scaledown is cheaper on a typical model: a median of $0.03 per 1M tokens (3:1 input/output) against $0.35 for Microsoft. The cheapest paid models are Microsoft: Phi 4 at $0.09 and ScaleDown: Extract at $0.03.
It depends on the job: both list 2 models, Scaledown the lower typical price, Scaledown the longest context (1M tokens). The workload table above shows which wins for chat, RAG, agents, batch jobs and long documents.
Neither — both list 2 models.
Scaledown: Microsoft's longest is WizardLM-2 8x22B at 66K tokens; Scaledown's is ScaleDown: Extract at 1M.
Neither lists a free model; both are pay-per-token.
Microsoft: 0 of 2 models accept images. Scaledown: 2 of 2.