3 ms·
This is Microsoft subsidizing Claude inference costs -- if you look at how they charge models against your allotment, Gemini, GPT-5 and Claude 4 Sonnet all cost
by cwyers 1y ago
This is Microsoft subsidizing Claude inference costs -- if you look at how they charge models against your allotment, Gemini, GPT-5 and Claude 4 Sonnet all cost the same, despite Claude 4 Sonnet being more expensive than the other two. Not really sure I understand the economics here, especially since there's not really a clear winner between GPT-5 and Claude 4 Sonnet for coding (if anything I think GPT-5 puts up a better showing).
- mattalex 1y agoIt might be that they pay less for anthropic depending how many tokens are generated by each model: total cost is token cost times number of tokens. I haven't checked gpt5, but it is not impossible that price wise they might be very comparable if you account for reasoning tokens used.
- poslathian 1y agoIs it possible that regardless of what they pay they think Anthropic is negative margin on it?
- martinald 1y ago"Sonnet being more expensive than the other two" -> you mean based on public pricing? Microsoft will not be paying retail prices for this.
- cwyers 1y agoI'm sure they don't pay retail on GPT-5 either.
- drewbitt 1y ago> I think GPT-5 puts up a better showing Would the more casual Copilot audience be OK with gpt-5-high - the model that many say is better than Sonnet - taking significantly longer to respond? Potentially minutes longer. A faster model can make sense as a default
- adonese 1y agoI think copilot is very aggressive on tokens and context size. That is how I guess the economy works for them.
- PhantomHour 1y ago> Not really sure I understand the economics here There is nothing to understand. The point of such subsidies is to turn OPEX into a green line on the stock market. Especially as Microsoft is currently also in a fight with OpenAI.