Release Notes
Added
max_tokens: sets the output ceiling of an Anthropic-dialect model. The
key goes on a models: entry or an agents: entry, and the agent's wins
over the model's, as for effort, temperature and top_p. It must be 1 or
more (config max_tokens 0: want 1 or more). A new session and a /model
switch evaluate it, the session records it in meta.json with where it came
from, and iota resume restores what was recorded — the default included —
whatever the config says today; a bot takes it from its config each start.
Thinking counts toward it. There is no /model tab or flag for it, and a
hand-edited max_tokens: 0 in a session is ignored. On any
other provider type the key does nothing and says so
(Warning: `max_tokens` does not apply to provider type … (ignored)).- A chat can move to another provider from
/model. A row on another
endpoint used to be refused with "start a new session there". Before the
first reply, picking it now moves the chat there. After a reply, iota first
saves the chat, then tells you how many earlier replies carry
provider-specific payload (reasoning, provider-side tool state) the target
does not replay, that iota resume <id> continues on the new model
afterwards, and that picking the old row moves it back — and asks
Move this chat to <provider>:<model>? before anything changes. Your
settings come along as they do on a -M resume (below). A chat that is not
saved is told to /save first, a model of another kind (an image model for
a chat) is still refused, and a bot never moves: its config does.
Changed
- The Anthropic dialect asks for up to 32000 output tokens, not 4096. A
thinking model could spend all of 4096 thinking and be cut off before it
wrote a word. The ceiling is a bound, not a charge: tokens are billed as
they are produced. An endpoint that caps lower answers a too-large value
with an error, so set a smaller max_tokens: for such a model.
- An output the ceiling cut off is an error, never the answer. On the
Anthropic dialect (stop_reason: max_tokens) and on the Responses dialect
(response.incomplete: max_output_tokens), the chat shows a red
Output truncated block: what cut it, the tool call it was composing if
any (dropped, by name — it does not run), and what to do. On Anthropic
that is where to raise max_tokens:, and that the edit reaches this
session only after a restart, a resume and a /model switch away and
back, while a lower effort in /model applies at once; the Responses
dialect has no such setting, so there it is to lower the effort in
/model or ask for a shorter reply. Text the model did write is kept,
marked incomplete; with no text or completed tool rounds to keep, the turn
is rolled back, though what the request cost still counts in the session
totals. A cut is not retried — the same request would hit the same
ceiling. A half-finished thought is no longer
shown, saved or sent back to the model as the reply. With -m, a cut fails
the run (exit 1, Error: output truncated — …, nothing saved), a cut-off
compaction summary fails instead of replacing the history with a fragment,
and a cut-off title leaves the placeholder title.
- Sessions record their provider by name, and resume there. A session now
stores the providers: name it runs on (session format v2) instead of a
provider type and a URL, and iota resume — and /session — opens that
provider from the config, as if -M <provider>:<model> had been typed. A
name the config no longer defines is an error that says how to resume it
elsewhere (… which the config does not define; resume it elsewhere with -M <provider>:<model>), never a quiet fall back to another endpoint, and
a missing key on the agent's own endpoint no longer stops a session that
runs on another. Resuming replies another dialect wrote says how many carry
payload this one does not replay. Sessions written by earlier versions
(format v1) are no longer read: opening one is an error that names
scripts/session-v1-to-v2.py, which converts them once. The script only
reports what it would change until you add --write; it rewrites
meta.json alone, leaving the transcript and attachments as they are. It
picks the provider by the endpoint the session recorded: the one entry of
that type whose url: is that URL, or the type's built-in provider when
the session ran on the default endpoint. When that does not settle it (no
entry matches, two do, or another type's entry has taken the built-in's
name), it reports the bundle and leaves it as it is for --map TYPE=NAME
to decide.
iota resume <id> -M moves the session for good. -M on a resume used
to change only that run; the session now records the new provider and
model, so the next plain iota resume continues there. What you set by
hand — effort, temperature, top_p, the output ceiling — comes along unless
the config there sets it; what the old endpoint's config supplied stays
behind. The warning says so./model rows read the same wherever the chat is. A row is
provider:model, or just the model when every choice is on one provider,
followed by the models: entries that name it in brackets
(relay:vendor/fast [fast]). What picking it does is an annotation beside
it, like (current): (moves the chat), or (needs a new session) in a
bot.- Anthropic calls without tools stream too. A
-m run with no tools, the
compaction summary and the title used a non-streaming request, which sends
nothing until the whole output exists and could outlast the 120 s wait for
a response. They now use the stream the chat uses, so an
Anthropic-compatible endpoint that cannot stream now fails there instead of
answering.
Fixed
- An Anthropic stream that stops early is an error, not an answer. A
response that ended before the message did — no message_stop, no stop
reason — was taken as complete, so -m and a compaction could succeed on a
fragment.
- A failing request without tools is sent 11 times, not 121. A turn
that offered the model no tools retried inside a retry, so one failing
request (a 5xx, a dropped connection) could go out up to 121 times with
minutes of backoff. Every model call now has one set of ten retries.
- Typing a choice's name in
/model picks that choice. Typing fast, the
name of a models: entry, sent fast to the provider as a model id; the
field now reads typed text as -M does, and provider:* picks nothing.
Install iota 0.7.0
Install prebuilt binaries via shell script
curl --proto '=https' --tlsv1.2 -LsSf https://github.com/iotash/iota/releases/download/v0.7.0/iota-installer.sh | sh
Install prebuilt binaries via powershell script
powershell -ExecutionPolicy Bypass -c "irm https://github.com/iotash/iota/releases/download/v0.7.0/iota-installer.ps1 | iex"
Install prebuilt binaries via Homebrew
brew install iotash/tap/iota
Download iota 0.7.0
